Skip to main content

About this book

This Conference proceeding presents high-quality peer-reviewed papers from the International Conference on Electronics, Biomedical Engineering, and Health Informatics (ICEBEHI) 2020 held at Surabaya, Indonesia. The contents are broadly divided into three parts: (i) Electronics, (ii) Biomedical Engineering, and (iii) Health Informatics. The major focus is on emerging technologies and their applications in the domain of biomedical engineering. It includes papers based on original theoretical, practical, and experimental simulations, development, applications, measurements, and testing. Featuring the latest advances in the field of biomedical engineering applications, this book serves as a definitive reference resource for researchers, professors, and practitioners interested in exploring advanced techniques in the field of electronics, biomedical engineering, and health informatics. The applications and solutions discussed here provide excellent reference material for future product development.

Table of Contents


PoratRank to Improve Performance Recommendation System

The e-commerce recommendation system has experienced tremendous progress and has caused an explosion of information, making it difficult for users to choose items that fit their preferences and take a long time. One way to overcome this condition is to use a collaborative filtering approach. Collaborative filtering generally uses similarity measurements and ranking predictions to produce recommendations. However, the recommendations presented are less reliable when data conditions are sparse. This condition encourages the development of ranking-based collaborative filtering. Some ranking-based methods are Copeland and Borda, which carry out an aggregation process to produce product ratings that are recommended to users. Both of these methods use limited ranking data at the user preference profile stage and do not involve re-ranking data during the aggregation process. This process causes the resulting recommendations to decrease in quality. Therefore, this paper proposes the PoratRank method. The basic idea of this method is to optimize the utilization of ranking data to produce product ratings that are more in line with user preferences. Ranking data is used as an additional factor in determining product points. Determination of product points not only looks at the ranking value but also considers the same number of ratings, and the position of the product in its appearance. It also sees the effect of the ranking value using the minus function. Optimizing ranking data in the aggregation process can improve the recommendation results, as shown by the average value (NDCG) of the PoratRank method, which is higher than the Borda and Copeland methods. The PoratRank method is faster than the Copeland method and manages to overcome the problem of sparsity and scalability, which is a major problem in the collaborative filtering approach, so the PoratRank method is feasible to be used in improving performance recommendations system.

Sri Lestari, Rio Kurniawan, Deppi Linda

Design and Implementation of a Registration System with Mobile Application at Public Health Center Based on IoT Using a RESTful API

In the health service center, many people often meet in a queue and tend to be chaotic, causing uncomfortable conditions and time inefficiency. From these problems, it was thought that technology was needed to make it easier for users to register and monitor queues. In previous studies, we have successfully implemented a digital queue system using a mini PC as the main control. This system has been implemented at the public health center “Puskesmas” Bojong Soang, Bandung. However, the mechanism of queuing, registration, or ticketing and monitoring could not be done online in an application. Therefore, in this study, an Android-based application was developed to register and monitor queues online and in real-time. The main purpose of this study is to build an integrated system between existing queuing machines and mobile applications so that the queuing mechanism can be done online. The proposed system consisted of three main components, namely the digital queue machine, server-cloud, and client application in the user’s smartphone. All modules were connected via the cloud with the developed protocol. This application was called “Q-Puskesmas”, which could run on the Android mobile platform with a minimum version of 5.0. The proposed system is able to display location identity, the latest queue number, and a menu for registering. From the test results, the application can update the queue number in real-time according to the existing conditions with a delay of <150 ms and also print a virtual registration number. This developed application was expected to facilitate the public in using health facilities/services.

Sugondo Hadiyoso, Akhmad Alfaruq, Yuli Sun Hariyani, Achmad Rizal, Tengku Ahmad Riza

Fuel Truck Tracking for Real-Time Monitoring System Using GPS and Raspberry-Pi

Trucks are the main transportation in fuel distribution from the dispatch center to the gas station. The lack of a security system of fuel shipments from the dispatch center to the gas station is an opportunity for theft by illegally moving fuel from trucks. Therefore, an application is needed to support the security system on the truck. In this study, we designed a fuel truck monitoring system by combining a global positioning system (GPS) module and a mini-computer. The purpose of this study is to implement an integrated which perform the function of tracking the position of the truck and to ensure the truck arrives at the gasoline station. The limit sensor on the tap will detect if the valve is open. This function is to ensure that the tap is only opened at locations with predetermined coordinates. Monitoring is done through online and real-time websites. The website will display a database of GPS where the coordinates of the fuel truck are located. If a theft is detected, the system will display a warning with the coordinates of its location. From the test results obtained tracking accuracy of 100%. The average difference in the coordinates of GPS tagging with GPS tracking of 5.48 m. The averaged of one-way delay and response time of the proposed system are 4.95 s and 160 ms respectively, it has good criteria according to the ITU-T G10.10 standard for real-time application. This designed system is expected to be implemented on existing fuel trucks for security reasons.

Rohmat Tulloh, Dadan Nur Ramadan, Sugondo Hadiyoso, Rohmattullah, Zikra Rahmana

Room Searching Robot Based on Door Detection and Room Number Recognition for Automatic Target Shooter Robot Application

Parameters of room are the presence of a door and room number, where to detected room need a system that can detected part of room. From it the researcher created a robot that can detected room target based on door and room number recognition. The way robot in finding a room is by tracing the corridor of the room by using the PID control method, PID value Kp = 5, Ki = 2 and Kd = 0.4, from it, the robot movement is stable with value is 55 RPM. To find room target the robot’s way of detecting a room is by recognizing the door frame, which is first processed using the Hough-transform method, where the results of it will be eventually processed as a parameter of a room door. After it the system will match the corresponding image to the existing image storage using template matching. After the door detection, the system will capture the image of the room, and process using OCR method and template matching. The experiment of robots in recognizing room doors has a success rate of 81.8%, while the success rate of robots in detecting room numbers is 84.6%, this result is due to the change of illumination level in the corridor of a room, and the size of the room number that affects the robot’s recognition. The result of experiment that has been done, the robot has a 88.8% success rate in finding the room, and has a time of 43–65 s from the robot’s starting point. From result, the researcher expected to be applied to the shooter robot application.

Syahri Muharom, Ilmiatul Masfufiah, Djoko Purwanto, Ronny Mardiyanto, Budi Prasetyo, Saiful Asnawi

A Comparative Evaluation of Acceleration and Jerk in Human Activity Recognition Using Machine Learning Techniques

Acceleration data have been widely used to study human activity recognition. However, the acceleration data collected from the accelerometer do not consider the force of gravity. Thus, it has a difficulty in discriminating closely similar activities. Jerk, the derivative of acceleration, is able to describe the changes of body accelerations correctly and mutually exclusive from the sensor orientation. This study aimed to compare the performance of jerk compared to acceleration data. The dataset used to do the comparison was collected from a triaxial accelerometer built-in Samsung smartphone with a sampling frequency of 50 Hz attached to the waist of thirty subjects. The subjects performed walking on a flat surface, walking upstairs, walking downstairs, sitting, standing, and lying down. The feature data were then grouped into three categories: acceleration features, jerk features, and combined features of acceleration and jerk. The evaluation was done using k-Nearest neighbors (KNN), linear discriminant analysis (LDA), support vector machine (SVM), and random forest (RF). The results showed that jerk features performed worse than the acceleration features. However, the combined acceleration and jerk features yielded the highest accuracy with above 87% for all classifiers. The present findings show that acceleration is still better than jerk in recognizing the pattern in human activity. The present study is, therefore, increasing the understanding of acceleration data and its derivative in human activity recognition.

Nurul Retno Nurwulan, Gjergji Selamaj

Rapid Thermal Modelling of Power MOSFET Using Bump Test Method to Evaluate Performance of Low Cost Heatsink

Heat dissipator or heatsink is crucial for the reliability of MOSFET-based converters and inverters. Currently, products such as To-220 are low cost passive aluminium heatsink that fit to a single discrete MOSFET package did not followed by proper documentation about its thermal model and performance. Thermal modelling has been used widely for thermal studies, but the methods were often time consuming, complex setup, and require sophisticated equipments. This study proposed a different method in thermal modelling that is specifically applied to evaluate the performance of low cost heatsink through the Bump Test experiment. It emphasizes on rapid, simple setup and calculations. The experiment were finished in less than 1000 s in order to obtain dataset that were used to develop case to ambient thermal model of IRF840 n-channel MOSFET that was constructed by third order transfer function. Also, it was converted to Cauer equivalent RCs circuit to obtain physical meaning of heatsink block layers. The conducted experiments were applied to two different low cost heatsinks shapes; heatsink A resulted in a dissipation of MOSFET’s power limit to 4.21 W, while the attached heatsink B produced 5.38 W. In other words, heatsink B has 27.79% higher dissipated power handling, or 13.06% higher continuous Drain current utilization, than heatsink A. These values represented quantitative measure of heatsink performance. This study can be used as a technical reference for MOSFET based electronic circuit board fabricator in order to obtain rapid measure of low cost heatsink performance.

Dista Yoel Tadeus, Heru Winarno, Priyo Sasmoko

Facial Skin Type Classification Based on Microscopic Images Using Convolutional Neural Network (CNN)

Skin is part of the human body that has a function as a barrier from the external environment and gives a physical appearance to an individual. In general, human skin types are classified into normal, dry, oily and combination skin. These skin types affected by the amount and change in facial sebum secretion and hydration ability. Determination of the type of facial skin is needed to determine skin care products and cosmetics in accordance with the type of facial skin they have. In this research, a system design for digital-based facial skin types using Convolutional Neural Network (CNN) method which has advantages to produce features and characteristics from the microscopic images dataset. The primary datasets taken directly using microscopic cameras and have been validated by a dermatologist. The CNN proposed model in this study consists of 3 hidden layers that use 3 × 3 size filters with output channels 8, 16 and 32 respectively, fully connected layer and softmax activation. The proposed model was able to classify the skin types into normal, dry, oily skin conditions and the combination with the best accuracy of 99.5% from 1200 training images and 400 test images used, meanwhile the parameters of recall precision and f-1 score produce values close to 1, which means that it is almost perfect or it can be said that the error is small.

Sofia Saidah, Yunendah Nur Fuadah, Fenty Alia, Nur Ibrahim, Rita Magdalena, Syamsul Rizal

Electrodermal Activity-Based Stress Measurement Using Continuous Deconvolution Analysis Method

Stress is a psychological response that must be handled properly, or it can affect psychology and trigger other complicated diseases. Nowadays, stress measured by filling out stress-related forms or interviews with a psychologist. This method still requires a willingness from the patient to consult their condition, which is often not giving the best result due to late treatment. Therefore, we need a system to monitor a person’s stress level conditions independently so they could be threatened quickly if their state gets worsen. One of the non-invasive methods that can be used to measure stress is an Electrodermal Activity (EDA) sensor. EDA is known to be associated with sweat gland activity regulated by SudoMotor Nerve Activity (SMNA), which response to a stressor. However, the SMNA need to decompose into the Skin Conductance Response (SCR) before it could represent stress condition. Hence, the decomposition stage is essential in measuring the quality of the stress condition. In this study, the EDA sensor placed in the palm to obtain the skin conductivity signal. Then, the skin conductivity signals were deconvolved using the Continuous Deconvolution Analysis (CDA) method to capture the SCR. CDA was chosen because it could produce a zero-based line SCR driver and could separate two overlapping peaks in SCR. The result, CDA is adequate to extract SCR, which is shown by the zero baseline of the SCR driver line with a clear peak. The clear SCR peaks are essential for identifying the responses related to the stimulus (stressor).

Yang Sa’ada Kamila Ariyansah Putri, Osmalina Nur Rahma, Nuzula Dwi Fajriaty, Alfian Pramudita Putra, Akif Rahmatillah, Khusnul Ain

Monitoring Stress Level Through EDA by Using Convex Optimization

This study designed an effective system for monitoring stress levels based on Electrodermal Activity (EDA) sensors by knowing the signal patterns and characteristics using convex optimization decomposition. The EDA sensor considered accurate and sensitive for identifying stress by analyzing the skin conductivity (SC) due to the changes in sympathetic nerve activity. However, the SC signal consists of phasic and tonic components, which needed to decompose to analyze stress levels. The SC signals also followed by the white Gaussian noise, which represents the error value. Hence, deconvolution is a crucial stage for the further process because the quality of the measurement depends on this result. This study aims to deconvolve the SC signal using the convex optimization method (cvxEDA). This model is physiology inspired by EDA based on Bayesian statistics, convex mathematical optimization, and sparsity. This research conducted with 18 subjects through three sessions of measurement. The given stimuli arise in each session to increase the level of stress. The results showed that this method could separate and identify SC. The Phasic component shows an increase in the stimulus of each session, as seen from the number of peaks. In contrast, there were no significant differences in the tonic component. This study shows that the phasic component is closely related to changes in sudomotor nerve activity (SMNA) and response to a stressor, which could be useful to classify stress levels in the future study.

Nuzula Dwi Fajriaty, Osmalina Nur Rahma, Yang Sa’ada Kamila Ariyansah Putri, Alfian Pramudita Putra, Akif Rahmatillah, Khusnul Ain

Design of Banana Ripeness Classification Device Based on Alcohol Level and Color Using a Hybrid Adaptive Neuro-Fuzzy Inference System Method

The ripeness classification device of banana-based on Internet of Things (IoT) was designed utilizing the Adaptive Neural Fuzzy Inference (ANFIS) method is presented. The ripeness of banana was identified from its color and alcohol level, utilizing MQ-3 gas sensor and TCS34725 color sensor. The built device has consisted of Arduino Mega, ANFIS for classification tool, and NodeMCU as the IoT Gateway. Google Firebase was used as an IoT platform and a database for the classification result. The inputs for the classification process were obtained from the alcohol level and RGB values from the sensors. The results showed that the device was successfully classified the ripeness level with an accuracy of 99.07%, that categorized into unripe, ripe, or rotten conditions. The final tuning of ANFIS classifier was using: 960 datasets and divided into 864 training data and 96 test data, Pi-shaped MF as input MF type, Linear MF as output MF type, (3-2-2-2) structure as the structure of MFs and Hybrid Algorithm as the optimizer. The average required time for predicting the ripeness level was 3.57 s. It is expected this device could be implemented for advanced applications such to classify other types of fruits and/or vegetables.

M. R. E. Ariono, F. Budiman, D. K. Silalahi

Design of Electrical Energy Storage System Produced by Thermoelectric Generator

Many home appliances can generate waste heat that are released to the surrounding. This waste heat can be utilized by converting it directly into electrical energy using a Thermoelectric Generator (TEG). However, the electrical energy produced by TEG from waste heat is quite low, hence a proper energy storage is essential to enable its use. This study aims to design an electrical energy storage system for TEG and assess its performance and capabilities. Six TEG module TEC1-12706 connected in series were utilized to harness the waste heat energy. A refrigerator compressor was used as the waste heat producer and a 3.7 V 400 mAh Li-Po battery was used as the electrical energy storage. A boost converter CE8301 was utilized to increase and stabilize the output voltage to extend the battery life. The storage system proposed in this study was successfully increased 27,3% of the battery capacity during 120 min of charging.

Rizky Septiawan, Mohamad Ramdhani, Wahmisari Priharti

Performance Comparison of Three Thermoelectric Generator Types for Waste Heat Recovery

Waste heat recovery using thermoelectric generator (TEG) is believed to be promising solution to the needs of the community’s electrical energy sources. TEG can produce electrical energy when there is a temperature difference between two different semi-conductor materials hence creating a voltage difference and a current flow. There are several types of TEG commercially available at the market with different specification and capabilities. This research was conducted to compare the power generation capability (Watt/m2) and the power production cost (Rp/Watt) of three TEG module i.e. TEC1-12706, TEC1-12710 and TEC1-12715. The test was conducted with and without the cooling system for the TEG module. The results showed that TEC1-12706 provide the best performance with 3800 W/m2 and 12 $/Watt with heatsink and 475 W/m2 and 97 $/Watt without heatsink.

Adhitia Rachman, Wahmisari Priharti, Mohamad Ramdhani

Cross-Gender and Age Speech Conversion Using Hidden Markov Model Based on Cepstral Coefficients Conversion

Animation movies often use children’s characters and they need children aged 5–10 to do a dubbing. For cost efficiency, a speech conversion can be done to support dubbing a children’s speech. To deal with it, in this research we propose the method to converting an adult’s speech to children’s speech. The contribution of this study is to design a signal processing algorithm to perform the conversion. In this study we propose a conversion method using the Hidden Markov Model (HMM) based on Cepstral Coefficients Conversion. The input is the speech of source speakers and the target speakers that using similar sentences. Features extraction, which is used is by extracted pitch (f0) and cepstral in conversion process, and the modeling method is HMM. System output is converted speech signals that has similar characteristics with target speech signal. From the testing results, the most optimal HMM parameter is using 4-state. The highest increase of cepstral Root Mean Square Error (RMSE) before conversion and after conversion is equal to 32.35% and an average 25.83% which obtained from 400 samples. Mean Opinion Score (MOS) on a scale from 1 (converted speech is very dissimilar with the target speech) to 5 (converted speech is very similar with the target speech). It resulted an average value of 2.505 in terms of similarities and has an average value of 2.805 in terms of quality which obtained from 30 respondents. The proposed method is expected to be used in the animation film industry in order to simplify and make efficient the dubbing process.

Meisi Aristia H. Gultom, Raditiana Patmasari, Inung Wijayanto, Sugondo Hadiyoso

Design of Electric Wheelchair with Joystick Controller as Personal Mobility for Disabled Person

The movement aids for people with motor system disability, especially the legs, has been currently very essential, because they also surely want to move freely from a place to another without bothering their family or closest friends. Therefore, this electric wheelchair research was made to facilitate the movement of people with disabilities within their activities, such as going to certain places or doing something. The design of the electric wheelchair was arranged in such ways to assist the body movements using joystick control connected to microcontroller Arduino. If the joystick lever was directed forward, backward, left and right, then the joystick command would be processed by the Arduino, so that it provided voltage to the motor driver, and the wheelchair could move forward, backward, turn left, and turn right. From 25 joystick command tests, it was shown the forward accuracy was 88%, the backward was 100%, the left turn was 100%, and the right turn was also 100%. Based on the test that have been conducted, the accuracy values represented that this electric wheelchair using joystick controller can be implemented as its function.

Hanifah Rahmi Fajrin, Thony Ary Zain, Muhammad Irfan

Comparison of the Effects of Feature Selection and Tree-Based Ensemble Machine Learning for Sentiment Analysis on Indonesian YouTube Comments

The main problems in sentiment analysis models on Indonesian YouTube comments are unstructured data and low classification accuracy. Sentiment analysis for Indonesian, which is different from English, requires proper preprocessing and classification methods. Previous research usually using Linear Support Vector Machine (SVM), Naïve Bayes and Decision Tree. Although the accuracy of SVM is better than other algorithms, it still needs to be improved. This study aims to compare the performance of the tree-based ensemble method and feature selection to improve the sentiment analysis model for Indonesian YouTube comments. This research crawled Indonesian YouTube comments from different domains and produce ten datasets. The preprocessing’s method in this research was removed stopword, convert slang words, and stemming. For feature selection, we tested two vectorizer method, i.e. Term Frequency (TF) or Term Frequency/Inverse Document Frequency (TF-IDF). The model build using six machine learning, consist of four tree-based ensemble machine learning to raise better accuracy, Linear SVM and Decision Tree. We use tree-based ensemble machine learning, they are Random Forest, and Extra Tree represents bagging ensemble. AdaBoost and Gradient Boosting represent boosting ensemble. SVM and Decision tree as a comparison. Based on experiments by combining feature selection and ensemble machine learning, it can be concluded that the type of vectorizer has little effect on classification accuracy. In all experiments, the best machine learning methods are Extra Tree with an accuracy of 93.39% and AdaBoost with an accuracy of 92.53%. Whereas, the use of TF or TF-IDF does not significantly affect accuracy.

Siti Khomsah, Ahmad Fathan Hidayatullah, Agus Sasmito Aribowo

Assessing Beads Generation in Fabricating Nanofiber Bioactive Material-Based Associated with Its Fluid Factors

Nanofiber materials for biomedical applications inevitably need to control their morphology to reach their best performance. However, it has a challenge regarding occurring beads formation during the fabrication process. Even though beads shape has already observed in numerous nanofiber experiments, a deeper study regarding a bead formation particularly related to fluid factors in bioactive materials have not reported yet. The main objective of this report is to elaborate the beads development particularly in fabrication nanofiber bioactive material-based and to investigate the result associated with the fluid factors (viscosity, conductivity, and surface tension). In this research, bioactive material chitosan (positive charge) and pectin (negative charge) are mixed with PVA (an electrospinnable polymer) in assorted composition PVA/Chitosan-Pectin (v:v) specifically 90/10, 80/20, 70/30, and 60/40. Based on quantification, 90/10 had 15 beads and got 92.7±19.59 μm2 of average area. In 80/20 composition, it had 57 beads with 121.83±19.78 μm2 average area. For 70/30, it had 86 beads with 111.6±24.46 μm2. And 60/40 had 117 beads with 129.8±19.13 μm2 average area. The result showed that the bigger bioactive material added to system, the more beads are formed. Regarding fluid factors, viscosity found that it had an opposite correlation with beads formation. Meanwhile, the relationship between conductivity and surface tension were observed complemented associated with beads formation. In the bead precision perspective, interestingly, the non-uniformity of the bead shape had no relation with fluids factors proofed by its average area and standard deviation in respective composition.

Muhammad Yusro

Intelligent System of Handling In Vitro Fertilization (IVF) Patients Post Embryo Transfer to Reduce the Level of Patient Anxiety and Help Fertility Doctors Quickly Answer Patient Questions

77.7% of IVF patients have high anxiety and 83.3% of patients have failed IVF programs. The strong desire to have a biological child and the process of IVF that is not easy, makes patients protective of the IVF process that is being undertaken. If patients experience health complaints, then patients immediately communicate with fertility doctors through the existing communication media. However, the slow time doctors give answers to patients, is a cause of increased anxiety of patients. Doctors have a strong desire to be able to quickly answer questions from patients, but are constrained by other medical activities that are the responsibility of doctors. Research aims to find a smart system method that can help IVF patients control anxiety and help fertility doctors answer questions from IVF patients. Primary data triangulation to obtain valid data, strengthened by validity testing and reliability testing. Secondary data to obtain evidence of the accuracy of a case-based reasoning (CBR) smart system and the challenges of producing quality system outputs. Based on the results of primary and secondary data analysis, a CBR modification was carried out by adding the role of patient feedback as a quality controller for system output. Anova test resulted in the value of F = 9902 and the coefficient test resulted in a value of t = 3147. The result of the research is a modified CBR for a smart system that helps IVF patients control anxiety and helps fertility doctors answer IVF patient anxiety questions quickly and accurately.

Paminto Agung Christianto, Eko Sediyono, Irwan Sembiring, Sutarto Wijono

Object Detection for Using Mask in COVID-19 Pandemic with Faster R_CNN Inception V2 Algorithm

When this research was done, based on data from , the number of positive patients reached 127,083, the increasing number of COVID-19 patients in Indonesia, and the new normal period had been imposed, making people more concerned about the health and the danger of spreading COVID-19 virus. New normal means were starting a new habit that is the habit of washing your hands and use a mask. Everyday activities in normal times require us to work as usual, in an atmosphere with other people. Everyone is required to use a mask to support the policy of using a mask, a system that can detect whether someone is wearing a mask or not. This research aims to be able to identify the use of masks in public areas. The dataset used is 5000 images of people wearing masks. This research uses the Faster R_CNN Inception V2 Algorithm. And then the results were evaluated using COCO mAP Score yielding in 0.58 mAP, with 0.78 mAP for large objects, 0.65 mAP for medium objects, and 0.48 mAP shows this research can contribute to the community to care in the use of masks. The model that has created in this study is used to detect the use of masks in public areas by using an image captured by a camera mounted in a particular place.

Apri Junaidi, Jerry Lasama

Long Range Ultrasonic Testing System Based on Lamb Wave Method for Validation an Optimized Piezoelectric Sensor Gap Array

Non-Destructing Testing (NDT) is crucial in the oil and gas industry to avoid fatal pipeline accidents and alarming property damage. Its general applications include the detection of pipeline defects in nuclear power plants, steam generator tubing, aircraft etc. Meanwhile, the Piezoelectric with Lamb Wave method testing system (PZ-LW), also known as Long Range Ultrasonic Testing (LRUT) is one of the methods in NDT used for crack detection on long-distance surface area. The current research proposed the ultrasonic testing probe design for the ideal gap between Piezo sensor resulting in reflection signal optimization for Piezoelectric sensor array in carbon steel pipe inspection. Response Surface Methodology (RSM) used the signal wave frequency excitation and the distance between the sensors array were optimized for appropriate multiple defect detection on 60 mm diameter carbon steel pipe sample. An Artificial Axial and Hole defect was utilized to test its efficiency. The analysis of the experimental results suggests the accuracy of PZ-LW system to amount to 98.55% when determining the defect location (hole defect), above 96.05% (axial defect measurement), while 98% and beyond (identification of defect shape). The inspection position is achievable until 311,399 m on free damaged pipe, 64,377 m (axial damage) and 61,267 m (hole crack defect damage). The reflection signal on PZ-LW inspection indicated good feedback signal amplitude with high distance detection in axial crack defect. The comparison of experimental and simulation results in Simulation of Non-Destructed Testing (SIMNDT) has successfully proven the consistency and accuracy of the proposed PZ-LW system for carbon steel pipe inspection.

Kharudin Ali, Johnny Koh Siaw Paw, Damhuji Rifai, Nur Amalina Awang, Ahmed N. AbdAlla, Abdul Rahim Pazakadin, Chong Kok Hen

Fuzzy Logic Error Compensation Scheme for Eddy Current Testing Measurement on Mild Steel Superficial Crack

Accurate measurement the depth of defect is essential to ensure the reliability and integrity of the pipe line structure and safety of the personnel workers in the oil and gas industry. This study proposes intelligent algorithm base on Fuzzy logic scheme to compensate the depth error measurement of GMR sensor on mild steel superficial crack. Intelligent rules in Fuzzy logic scheme allows the propose method to be effective on compensate lift off effect in between 1 and 4 mm. The Eddy current testing probe is design by utilize the sensitive of GMR sensor on magnetic field in order to improve the accuracy depth crack measurement on mild steel. The Arduino Mega 2650 was used as a data processing device for receiving and transmitting of signal through the available fifty-four digital pins (as input and output) and sixteen analogue pins. MATLAB is use as a platform to design and implementation the Fuzzy error compensation. The experiment results show the proposed sensor error compensation scheme able to reduce the error measurement up to 23%.

Damhuji Rifai, Abdul Rahim Pazikadin, Kharudin Ali, Moneer A. Faraj, Noraznafulsima Khamsah, Ahmed N. Abdalla

Development of Cutting Force Measurement Instrument for Turning Tool Post Using Arduino UNO

Cutting force is one of the forces occur during turning process. A measurement instrument is necessary to monitor the cutting tools performance during a turning process operation. Nowadays, dynamometers employing piezoelectric sensors embedded into the machining system, but it very expensive. The strain gauge is a cheaper approach to develop the dynamometer. The purpose of this project is to develop a measurement instrument of cutting force for turning tool post using strain gauge. Cutting force signals were captured, amplified, conditioned, converted to digital signals and read by Arduino. The measurement instrument is calibrated via increment load from 5 to 40 kg. Then, the effects of forces with two difference conditions namely dry and flood fluid condition is performed during turning process. The corresponding force readings are obtained and displayed for real time monitoring. From the observation, it is found that there is a linear relation between force and strain gauge voltage. The mathematical modelling between these force and voltage is then developed. The recent study proved that the Arduino Uno can be used in developing cheaper force instrument measurement (dynamometer) for turning process.

Wan Zulkarnain Othman, Mohamad Redhwan Abd Aziz, Nor Hana Mamat, Ahmad Fikri Ramli

Spatial Mapping and Prediction of Diphtheria Risk in Surabaya, Indonesia, Using the Hierarchical Clustering Algorithm

Diphtheria cases in the city of Surabaya from 2015 to 2018 have increased every year. This disease can be prevented by immunizing DPT 1, DPT 2, and DPT 3 (Diphtheria, Tetanus Pertussis) given to school-age children. Immunization is the most dominant factor, where children who do not receive DPT immunization are five times more likely to be infected with diphtheria compared to children who are immunized. This paper proposes a new approach to diphtheria risk analysis in Surabaya based on multiple criteria, including DPT immunization, number of diphtheria sufferers, and population density using the hierarchical clustering algorithm. Information is presented in the form of spatial mapping of each urban village in Surabaya, Indonesia, thus, it can present information in a smaller scope. The hierarchy clustering average linkage algorithm achieves the smallest average variance value 3.3 × 10−5 and better than single linkage and complete linkage for 2016, 2017, and 2018. This developed application also provides a prediction of the next year’s diphtheria risk level. The results of the 2019 predictions show a better diphtheria risk level of vulnerability using a single linkage rather than average linkage and complete linkage with the smallest variance 4.43 × 10−5. which shows very good clustering results. The results of the knowledge shown in this application can be used as a decision support analysis for early vigilance in diphtheria in efforts to prevent and monitor diphtheria in Surabaya.

Arna Fariza, Habibatul Jalilah, Muarifin, Arif Basofi

Internet of Things in Flood Warning System: An Overview on the Hardware Implementation

Early warning in flood event can save property and life. The advancement of Internet of Things (IoT) technology along with cheap sensors makes IoT based flood warning system an attractive choice for disaster management. Various methods of IoT implementation have been used for the purpose of flood monitoring and warning. This paper presents an overview of literature related to the hardware implementation of IoT and the corresponding method of installation in flood warning system. The paper contributes by highlighting the sensors, microcontroller and wireless communication, IoT platform and method of installation employed in the literature adopting IoT in flood monitoring and flood warning systems. The paper further contributes by providing recommendations for the most suitable IoT hardware for a practical, cost effective and reliable flood warning system. Ultrasonic sensor that is largely used for water level detection needs to be waterproof to withstand the environmental elements. It is useful to include other sensors for measurement of various hydrological, meteorological and geological data for further use in flood prediction employing artificial intelligence and machine learning methods. NodeMCU ESP8266 is mostly used as it is a combination of microcontroller and Wi-Fi microchip. An update of NodeMCU ESP8266, the NodeMCU ESP32 is recommended as it is more flexible, more energy efficient and provides extra input output connectivity. LoRa which is a low power communication module that provides long range low data rate transmission, can be used for sensor nodes placed far apart providing a larger area coverage of flood warning system. As for IoT platforms, it is beneficial to use a combination of multiple platforms for widespread dissemination of flood alerts to ensure effectiveness of flood warning systems.

Nor Hana Mamat, Mohd Hafiz Othman, Wan Zulkarnain Othman, Mohamad Fadhil Md Noor

IoT-Based Solar Photovoltaic (PV) Real Time Monitoring System for Power Consumption on Maahad Tahfiz School

Electricity resource is one of importance thing needs in daily life, especially for urban and rural residents. Currently, the rural faces a big problem due to insufficient electricity resources which are completely dependent on the use of generators and thus affect the daily usage of electricity. To preventing on this problem, the Solar energy garners attention for its clean nature and renewability with regards to electricity generation. This paper presents an Internet of Things (IoT) monitoring system for real time power consumption on rural area especially at Maahad Tahfiz school in Terengganu. Some essential parameters on PV real time system include Temperature, Power, Anemometer and Irradiance which are sensed using the sensors. The system consists of Thinger cloud, data gateway, and smartphone with Android OS application display for actual value monitoring, indicating the relationship between irradiant and power produced of solar PV. The results demonstrated the sampling data collection stored in Thinger cloud in higher sampling data per hour (15 min/data) yielded higher data accuracy up to 80% as compared to the lower sampling data per hour (60 min/data), with 70% accuracy. the irradiant measurement is directly proportional to the power produced based on the data analysis on Response Surface Methodology (RSM). Henceforth, the total power produced per day for solar panel (PV) will be known accurately based on the online system development and at the same, the estimation time for power will be predicted based on the total capacity of power used in Maahad Tahfiz school in Terengganu.

Abdul Rahim Pazikadin, Kharudin Ali, Damhuji Rifai, Nur Amalina Awang, Ruzlaini Ghoni, Nor Hana Mamat

Auto Tracking Mobile Robot Navigation Based on HUE Color of Image Pattern

In industrial, a mobile robot is often used to move an object from its initial location to the final location. To perform the task, the robot needs to move forward following the path provided. The problem has been countered from movement the robot is to find a path that moves the robot from start position to goal while never touch any obstacle. For an auto-tracking mobile robot, the most critical part is the target identification and tracking of the moving object. Although the motion planning problem is defined in the regular world, it lives in another space. In this project, a combination of the HUE-based color of the pattern was used as a tracking object and the mobile robot can only identify the color of that image pattern for the auto-tracking navigation. Therefore, a study was carried out to make a path planning for mobile robot navigation using a vision system which is Pixy CMUcam5 as a sensory point to detect the color of the image pattern of the object for the auto-tracking purpose. Arduino Mega is the main control unit that controls all input and output of the mobile robot. The results of the work presented suggest that the color of image pattern identification used for the mobile robot navigation should be a combination should be at least two combinations so that, the mobile robot could differentiate between the tracking object and the environment.

M. Z. Muhammad Luqman, W. T. Wan Faizura, R. A. R. S. Nur Adiimah, A. R. Nazry, A. M. A. Hasib, A. G. A. Shahrizan

Implementation of EC and PH Value Monitoring for NFT-Based Hydroponic System Applying Internet of Things (IoT)

Nutrient Film Technique (NFT) is introduced as an uplifted technique in conventional hydroponics system. This soil free farming method offers an option to solve the global soil salinity issue. This technique is using fertilized water known as nutrient solution which flows through the plant roots. The nutrient concentration is indicated by the value of electrical conductivity (EC) and pH. An optimum value of EC is very important in producing healthy growing plants. Using conventional method, the reading of these two parameters manually. Thus, this will yield to ineffective data records and consume more time since the person need to attend at the plant area to take the readings. Therefore, through this research, an EC and pH sensor are incorporated in the NFT system to detect the parameter values. These values can be monitored and recorded in continuous real-time (online) using internet of things (IoT) through an android application-Blynk app. With less than 10% of data irregularities recorded, the research resulted that this system is reliable in monitoring real time data monitoring and recording used in NFT.

Lia Safiyah, Raja Siti Nur Adiimah, Farah Hanan, Suzanna, Khairul Irwan

State-of-the-Art Method to Detect R-Peak on Electrocardiogram Signal: A Review

The detection of the R peak on the ECG signal is very important to use to see the amount of variability in the heart rate. So that a person’s vital signs can be known if there are heart defects including arrhythmias. Besides, several recent studies suggest that the detection of the R peak in the ECG signal can also be used to detect respiratory rate signals. Therefore, an appropriate algorithm is needed to detect the R peak in the ECG signal so that there is no mistake in diagnosing a person’s physiological state. Several researchers have developed methods for detecting the R peak in the ECG signal with advantages and disadvantages. Therefore, this paper aims to provide a specific description of the R peak detection method to facilitate other authors in developing R peak detection methods in ECG signals. References in this paper are gathered from several journals and procurements regarding the definition of peak R. The results showed that some researchers used an adaptive threshold system with an accuracy rate of 99.41%, while some other researchers used the 99.5% decomposition method, some researchers used the convoluted neural network (CNN) method with an accuracy level of 99.7%. The benefit for the next researcher is to develop a real-time R peak detection tool to use to see the amount of variability in the heart rate by using one of these methods.

Anita Miftahul Maghfiroh, Syevana Dita Musvika, Levana Forra Wakidi, Lamidi Lamidi, Sumber Sumber, Muhmmad Ridha Mak’ruf, Andjar Pudji, Dyah Titisari

A Review on Robotic Hand Exoskeleton Devices: State-of-the-Art Method

An exoskeleton is a device that helps the process of medical rehabilitation for people who have disorders in using their limbs. A low cost, effective sensor, control system, and an actuator are still the central issue in developing exoskeleton devices. This study aims to review an exoskeleton device, development, and recent technologies. The contribution of this study is that the paper can be used as a guideline to design an exoskeleton device. Specifically, the focus of this review discusses hand exoskeleton design. This review discusses three things, namely control signal, control mechanism, and exoskeleton actuator. In terms of the control signal, it addresses several techniques to control the exoskeleton by utilizing EMG, EEG, voice, and FSR (forced sensor) signals. In terms of control mechanism, several studies utilize pattern recognition based on machine learning and virtual reality to assist in using the exoskeleton. In terms of the actuator, the exoskeleton that was designed still has some shortcomings, namely weight and ergonomic design. The review results show that EMG signals are more often used in controlling exoskeleton devices. In the method section, pattern recognition using machine learning is still a significant part of the development of exoskeleton. In the actuator section, DC motors and linear actuators are more widely used than other types of motors. So, overall, the exoskeleton can still be improved from various aspects to make the subject more comfortable in use.

T. Triwiyanto, Endro Yulianto, Muhammad Ridha Mak’ruf, Dyah Titisari, Triana Rahmawati, Sari Luthfiyah, Torib Hamzah, Syaifudin Syaifudin, I. Dewa Gede Hari Wisana

Experimental Validation of the Multifunctional Device for Measuring Forces and Torques on Spine Phantoms

In modern spinal surgery, minimally invasive procedures using mechatronic and robotic systems are actively developing. Moreover, for the development of optimal designs of such systems, information is needed on the range of forces arising from the interaction of medical instruments and patient tissues at all stages of surgical intervention. The authors have developed and manufactured a prototype of a multifunctional device for measuring forces and torques during operations. The purpose of this article is to describe the laboratory validation of this prototype. Two experiments of the multifunctional device were carried out on medical phantoms in conditions close to real medical practice. In particular, an imitation of the retraction of the back muscles was carried out, and the screw was screwed into a vertebral pedicle phantom. The data obtained made it possible to confirm the correspondence of the characteristics of the sensor elements of the device to the required measurement ranges. The operability of the device was proved, and the possibility of using the device in medical practice was confirmed. The results obtained will allow the transition to clinical trials of the prototype.

Mikhail A. Solovyev, Andrei A. Vorotnikov, Andrey A. Grin, Daniil D. Klimov, Yuri V. Poduraev, Vladimir V. Krylov

Advanced Biomechanical Systems: Development and Design of an Accelerometer-Based Prosthetic Sensorimotor Platform

Following stroke, injury, or exposure to physically limiting conditions, limbs can become physiologically compromised. In particular, motor and fine-dexterity tasks involving the arm, particularly in locomotion, grasp and release, can be influenced becoming either delayed and having to deal with greater force demands. Current prosthetic systems use electromyography (EMG)-based techniques for creating functional sensorimotor platforms. However, several limitations in practical use and signal detection have been identified in these systems. Accelerometer-based sensorimotor systems have been suggested to overcome these limitations but only proof-of-concept has been demonstrated. Here, we explore design specifications for accelerometers being developed for prosthetic integration. We have developed optimizations for the current model, evaluated system properties to enhance sensitivity and reduce signal noise, and performed a pilot test using simulation to test this model. The data suggest these novel design parameters can enhance signal detection, when compared to conventional accelerometers. Future avenues should focus on validation of this design prototype in a full prosthetic system.

Peter Anto Johnson, John Christy Johnson, Austin Mardon

A Simple Design of Sterilizer Equipment for Infant Incubator Using Ultraviolet Germicidal Lamps

Infant incubator is an electromedical device used to treatment for a newborn which has the possibility to become a place of spread infection. Therefore, this study aims to design a simple infant incubator sterilizer as an effort to reduce the level of nosocomial infection in the neonatal intensive care unit. The simple design of infant incubator sterilizer has been made using UV germicidal lamps, arduino mega, remote control, and equipped with a sterilization time setting. The device was tested under laboratory conditions, with observational parameters consist of sterilization time, a display showing the countdown time of sterilization and the hour meter of the ultraviolet lamp, as well as remote testing, to turn on/off the device. Furthermore, the effectiveness of sterilization was tested on three types of gram-positive bacteria namely Escherichia coli, Staphylococcus aureus, and Bacillus cereus with sterilization time variations of 30, 60, 90 and 120 min. The result of this research is a prototype of sterilizer equipment using an ultraviolet germicidal lamp that can be applied to infant incubators. The effectiveness of sterilization using this equipment showed a trend gradually decrease the number of bacteria.

Mamurotun, Indah Nursyamsi Handayani, Nur Hasanah Ahniar, Catharine Bernadette

Microstructure and Corrosion Behavior of Bioabsorbable Polymer Polylactic Acid-Polycaprolactone Reinforced by Magnesium-Zinc Alloy for Biomedical Application

Medical composites have a lot of applications in medicine and orthopedic. Biomedical/orthopedic composites belong to biomaterials (Biomaterial can substitute natural tissues in the body and perform their functions). In this research, PLA (Polylactic Acid) and PCL (Polycaprolactone) polymer matrices filled with Mg6Zn alloy and observed the microstructure and electrochemical impedance spectroscopy. PLA and PCL were combined with a ratio of 7:3, 6:4, and 5:5 then dissolved respectively using chloroform. After that, mixed with Mg6Zn with three separate compositions of 5, 10 and 15%, it was then mould into a petri dish until it was dry. The structure was observed using scanning electron microscopy (SEM JEOL, JSM-6390A Japan). The immersion test examined the corrosion rate by observing the change in the pH, weight loss, and solubility of the composite in the simulation body solution using UV-Vis by looking at the specific functional group wavelength absorption area. Uv-Vis Analysis was conducted using Metrohm auto lab spectroelectrochemistry. The The microstructure found that the PLA-PCL-Mg6Zn biocomposite content is homogeneously mixed and uniformly distributed. Corrosion activity found by the immersion test, the reduced mass is not so large that it will last longer for bioabsorbable biomedical applications. 5PLA5PCL10Mg6Zn is an optimal composition because it decreases with pH most a little from the other, although the weight loss is not the least insignificant.

Aprilia Erryani, Alfiyah Rahmah, Talitha Asmaria, Franciska Pramuji Lestari, Ika Kartika

Seizure Classification on Epileptic EEG Using IMF-Entropy and Support Vector Machine

Various methods have been developed by researchers to recognize brain abnormalities through EEG signals. One of the diseases or disorders of the brain is seizures in epilepsy. EEG signals in seizure conditions display a different pattern compared to EEG signals in normal conditions. Researchers analyzed the EEG signal using a variety of observed approaches. One phenomenon used to analyze EEG signals is signal complexity. Signal complexity captures fluctuating patterns of EEG signals quantizing them to distinguish normal and seizure signal conditions. In this study, we propose the proper feature extraction method based on the basic characteristic of the signal. We extract the EEG signal’s information using entropy calculation from the intrinsic mode function (IMF entropy). Our main goal is to distinguish normal and seizure EEG signals. The entropy is calculated from the IMF resulted from empirical mode decomposition (EMD), then entropy from the relative energy of each IMF. To test the performance of the proposed feature extraction method, the support vector machine (SVM) is used as a classifier. The highest accuracy is 86.3%, sensitivity is 86.33%, and the specificity is 93.17% for three data classes: normal, interictal, and seizure. The proposed method has the potential to improve its performance, considering there are still many variations of EMD methods and decomposition levels that can be evaluated. Furthermore, testing on more massive datasets is interesting to do in future research.

Achmad Rizal, Inung Wijayanto, Sugondo Hadiyoso

Influence of Thermo-Mechanical Processing on Microstructure, Mechanical Properties and Corrosion Behavior of Ti-6Al-6Mo Implant Alloy

In this work, titanium alloy with additional aluminum and molybdenum is expected to be used as a clip aneurysm material. Thermo-mechanical processing (TMP) is required to achieve excellent workability in this alloy. The TMP of Ti-6Al-6Mo alloy composition has been successfully prepared. The alloy was homogenized at 1100 °C–12 h. The method is then continued to solution treatment with T = 1100 °C–30 min and hot-rolled by 80% deformation. The last method is heat-treated with T = 1100 °C–1 h followed by quenching in different media of air, water, and ice. The corrosion behavior of this alloy was investigated using Hank’s solution by polarization method. In the Ti-6Al-6Mo alloy with a hardness value of 44.4 HRc, the TMP and the quenching in water obtained various nucleations of the α-phases, less of the lamellar α-phase and less of the α-colony. The water-quenched Ti-6Al-6Mo alloy has excellent passive film stability and is more likely to increase the alloy corrosion efficiency.

Ika Kartika, Fendy Rokhmanto, Yudi Nugraha Thaha, Ibrahim Purawiardi, I. Nyoman Gede Putrayasa Astawa, Aprilia Erryani, Talitha Asmaria

Covid-19 and Tuberculosis Classification Based on Chest X-Ray Using Convolutional Neural Network

The high rate of patients with tuberculosis (TB) with the graph showing a continual increase requires the research in any sector as the programs to eradicate tuberculosis. One of the applications is the Decision Support System (DSS) that helps the medical experts particularly doctors in diagnosing TB grade 1+. 2+ , and 3+ rapidly. Another problem is related to the imbalance between the number of patients and the number of medical practitioners in the condition of pandemic Corona Virus Disease (Covid-19) today. Hence, DSS is highly required and it can be used for the long-term management of Covid. In this study, the rapid classification of normal lung, tuberculosis lung, and Covid-19 lung based on the Chest X-Ray (CXR) image was proposed as the initial step of DSS implementation. The proposed image processing based CXR classification using Deep Learning Convolutional Neural Network (CNN) obtained the highest accuracy rate of 88.37%. This accuracy was obtained in the second scenario with the 208 CXR datasets. The small number of datasets used was related to the limited number of CXR Covid-19 images with good quality brightness. The proposed system developed is expected to help doctors in diagnose lung disease.

Suci Aulia, Sugondo Hadiyoso, Tati L. E. R. Mengko, Andriyan B. Suksmono

Control of Wheelchair on the Ramp Trajectory Using Bioelectric Impedance with Fuzzy-PID Controller

A person who has a severe disability, have difficulty walking, climbing, and descending stairs, so requires assistive devices for daily mobility. in particular, the paralysis of the upper and lower limbs resulted in them being unable to use the wheelchair properly. So that in this study, an electric wheelchair that can be used by persons with disabilities is made, which is able to move based on the bioelectrical impedance signal on the part of the body that can still be moved. The pre-existing research is controlling wheelchairs on flat roads, whereas when the road goes up or down, the wheelchair has an unstable speed. This research was developed by utilizing Fuzzy-PID control, which is able to control the wheelchair speed so that it is more stable. This electric wheelchair with Fuzzy-PID control is able to walk on an up and down ramp at an angle not exceeding 10°, at a stable speed. The trial subjects used were four people with different weights. The error rate at each speed increases as the ramp gets steeper, and the user load gets higher. Success in this trial is at an average success rate of 80%. This occurs because of the change in the width of the bioimpedance signal, which affects the slow and fast commands when given the command to move, and the tilt angle of the track which affects the motor speed when the wheelchair crosses the track.

Masyitah Aulia, Achmad Arifin, Djoko Purwanto

Unstable Walking Detection in Healthy Young Adults Using Postural Stability Index

Accelerometer is a common device being used in healthcare industry to recognize human daily activity as the early detection for the risk of fall. Past studies have been carried out in order to develop an effective human activity recognition using an accelerometer, either accelerometer in the form of dedicated motion sensors or the built-in triaxial accelerometer in smartphones. This study aimed to utilize the postural stability index (PSI) in discriminating the subtle changes in closely similar daily activities, such as walking on a flat surface, walking upstairs, and walking downstairs. The walking activities above were selected to see whether the PSI could detect the slight postural sway in the acceleration data. The dataset was collected from twenty-four subjects with a smartphone-based triaxial accelerometer inside the subjects’ front pocket. The ensemble empirical mode decomposition (EEMD) was used to pre-processed the collected data. Then, the PSI values were obtained by calculating the complexity index of multiscale entropy (MSE) for each intrinsic mode function (IMF). The data calculation using paired t-test showed a significant difference between walking downstairs and walking on a flat surface (p = 0.039 and p = 0.02 for 2-tailed and 1-tailed t-test, respectively). The PSI value of walking downstairs was higher than walking on a flat surface. From this result, it is evident that walking downstairs is more challenging and less stable than normal walking. In conclusion, the PSI could be used for the early detection of unstable walking.

Nurul Retno Nurwulan

Computational Study of Ventral Ankle-Foot Orthoses During Stance Phase for Post-surgery Spinal Tuberculosis Rehabilitation

Spinal Tuberculosis could cause abnormalities in the lower extremity, even after the surgery and can be treated with Ankle-Foot Orthoses (AFO). The patient with that condition has to stabilize their body because it tends to fall forward. Thus, a ventral AFO is needed to support them. This study aims to evaluate AFO design by using Finite Element Method (FEM) with two different thickness (2 and 4 mm). The material used was a Polyethylene Terephthalate (PET). Three subphases in the stance phase were simulated, namely initial contact, midstance, and terminal stance. The external force in the simulation was based on the patient’s weight (50 kg). The result showed that the highest stress was obtained from ventral AFO with a thickness of 2 mm at the terminal stance, 244.7 MPa. This value exceeded the Ultimate Tensile Strength (UTS) of PET which indicates the design would break. The ventral AFO with a thickness of 4 mm had lower stress than the UTS of PET. The strain of both designs was quite low with the maximum value of 0.019 from the ventral AFO with a thickness of 4 mm in the terminal stance. The deformation was also acceptable for both designs with the highest value of 39.67 mm from the ventral AFO with a thickness of 2 mm in the midstance. In conclusion, the ventral AFO with a thickness of 4 mm could be used for post-surgery spinal tuberculosis patient’s rehabilitation.

Alfian Pramudita Putra, Akif Rahmatillah, Pujiyanto, Khusnul Ain, Nur Khafidotur Rodiyah, I. Putu Alit Pawana, Lolita Hapsari Dwi Syahananta, Mohammad Rizki Dwiatma, Arief Sofian Hidayat

The Implementation of EEG Transfer Learning Method Using Integrated Selection for Motor Imagery Signal

Brain-computer interface (BCI) is a system that can translate, manage, and recognize human brain activity. One of the devices used in the BCI system is Electroencephalogram (EEG). The brain signals produced by the EEG are diverse. One of them is the motor imagery signal. Motor imagery signal is used to translate the EEG signal into a specific movement. The performance of motor imagery signal classification is influenced by the number of training and testing data used. In most cases, the training data consists of a higher number of trials than the testing data. However, more trials cause higher subject variation. Previously study mentioned that this problem can be overcome by using transfer learning methods, which aimed at simplifying the training model. In this study, transfer learning in BCI is implemented using the integrated selection (IS) method, which simplifies the training model. Furthermore, IS is optimizing the data by removing the irrelevant channels of the EEG signals. Integrated selection uses the CUR matrix decomposition algorithm. The method split the data into two components, namely identity and historical data, represented by the C and UR matrix, respectively. The characteristic of the data from IS then calculated using three feature extraction methods. They are Fast Fourier Transform (FFT), Hjorth Descriptor, and Common Spatial Pattern (CSP). The features are then classified using the k-Nearest Neighbor (K-NN) method. The use of IS in the BCI system increases the accuracy of more than 6% and six-times faster processing time. In general, the integrated selection method is able to improve the performance of the BCI system.

Aris Ramadhani, Hilman Fauzi, Inung Wijayanto, Achmad Rizal, Mohd Ibrahim Shapiai

Computer Aided Diagnosis for Early Detection of Glaucoma Using Convolutional Neural Network (CNN)

Glaucoma is an eye disease caused by an increase in eye pressure resulting in damage to the optic nerve and can cause blindness. An ophthalmologist makes the diagnosis of glaucoma by analyzing the fundus images that show the retinal structure. This manual diagnosis requires years of expertise and prone to error. Previous studies have designed a glaucoma CAD system based on Convolutional Neural Network (CNN) and showed promising results. This study proposes the CNN method consisting of three hidden layers that use 3 × 3 of the filter size with 16, 32, 64 output channels, fully connected layers, and sigmoid activation. The experiment is conducted using the RIMONE R2 fundus images dataset to classify normal and glaucoma conditions. From 455 fundus images, 75% are used as the training data, while the rest is used as the validation data. From the experiment, this study outperforms other previous studies by achieving 91.22% of accuracy. The glaucoma system detection that has been developed in this research, can be helpful for ophthalmologists to establish an initial diagnosis of glaucoma that can reduce the harmful effects of glaucoma.

Yunendah Nur Fu’adah, Sofia Sa’idah, Inung Wijayanto, Nur Ibrahim, Syamsul Rizal, Rita Magdalena

Invasive Ductal Carcinoma (IDC) Classification Based on Breast Histopathology Images Using Convolutional Neural Network

Invasive Ductal Carcinoma (IDC) is the most common sub-type of all breast cancers that affected adult women worldwide. IDC can spread to other areas of the body such as liver, lungs and even bones. The process of identifying and categorizing breast cancer sub-types accurately is a very important clinical task. IDC Diagnosis requires extremely serious measures, such as surgery and radiation therapy. Diagnosis based on pathological imagery is no less difficult, requiring a microscope and manual learning to classify it as positive or negative cancer. This process is very time consuming and conveys many errors due to human cognitive limitations. The existence of a system which can automatically perform such work, is expected to save time and reduce the error rate diagnose. This study proposed IDC and Non IDC classification by analyzing the Breast Histopathology Images using Convolutional Neural Network (CNN) method. The dataset consisted of 1020 IDC images, the same number is also used for Non-IDC. The model was composed of CNN with three hidden layers plus one fully connected layer with sigmoid activation. An evaluation is carried out to see the performance of the proposed method by using a matrix of precision, recall, F1, and accuracy. The experimental results show that the proposed method provides precision, recall, F1-score of 0.93 and 93% accuracy. This study is expected to be validated for later use in assisting medical authorities for conducting clinical diagnoses.

Nor Kumalasari Caecar Pratiwi, Yunendah Nur Fu’adah, Nur Ibrahim, Syamsul Rizal, Sofia Saidah

Drowsiness Detection Based on EEG Signal Using Discrete Wavelet Transform (DWT) and K-Nearest Neighbors (K-NN) Methods

Drowsiness generally occurs due to lack of sleep. Drowsiness can trigger various problems, such as decreasing productivity, damaging emotions, even to the point of causing serious accidents, both on the highway or in the workplace environment. One possible way to detect drowsiness is by using an Electroencephalographic (EEG) signal. EEG is a test used to evaluate the electrical activity in the brain. This research proposed a system that can detect drowsiness based on EEG signal using Discrete Wavelet Transform (DWT) as feature extraction and K-Nearest Neighbor (K-NN) as classification method of drowsy and normal conditions. At a preliminary stage, the system would perform a pre-processing to minimize noise signals using normalization and grounding magnitude. Feature extraction of these EEG signals was then decomposed using DWT function whereas the K-NN method is used to classify the EEG signals either in normal or drowsy conditions. The K-NN is done by Euclidean Distance Method. The private dataset consists of 60 signals, divided into 30 signals to normal and drowsy each. This research used DWT with eight-level decomposition of Alpha and Beta signals, and 3 wavelet family types (Coiflet, Symlet and Biorthogonal). Based on the results of tests conducted, EEG signals was decomposed using 3 different types of wavelet family generally provides accuracy values that are not much of a difference while selecting different K values for K-NN classification affects the accuracy. In conclusion, the value of k = 5 is the optimum value to classify normal dan drowsy condition. This condition is in accordance with the K-NN theory in which a greater k value can reduce noise in the classification process so it can improve accuracy of the system. This condition provides system performance with the highest accuracy around 90–100% for any type of wavelet family.

Cahyantari Ekaputri, Yunendah Nur Fu’adah, Nor Kumalasari Caecar Pratiwi, Achmad Rizal, Alva Nurvina Sularso

Classification of White Blood Cell Abnormalities for Early Detection of Myeloproliferative Neoplasms Syndrome Using Backpropagation

One of the diseases of white blood cells is myeloproliferative neoplasms syndrome where this disease is an abnormality in the bone marrow in excessive blood cell counts. Full Blood Count (FBC) is a type of examination that shows the patient’s health status, blood cell abnormalities, and the presence of infection in the patient’s body. However, the determination of cell abnormalities is still done manually based on the knowledge and experience of clinical pathology so that the determination of these abnormalities is subjective. Therefore we need a system that is able to classify white blood cell abnormalities automatically, objectively and accurately. This study uses digital image processing on peripheral blood smear images then feature extraction will be obtained which will be the input of the classification system. The features used are area, perimeter, metric and compactness, while the classification method used is the backpropagation method. The best backpropagation network architecture used is 4, 6, 8, 5, 2 with a variation of learning rates of 0.05 and 0.3 producing the best accuracy rate of 91.82% with the amount of training data that is 516 and testing data is 159.

Zilvanhisna Emka Fitri, Arizal Mujibtamala Nanda Imron

Non-Complex CNN Models for Colorectal Cancer (CRC) Classification Based on Histological Images

Colorectal cancer become a significant public health issue and is the world’s second leading cause of death from cancer. Cancer becomes a very dangerous disease, because it gives no visible signs at an early stage. Signs of cancer will usually only be seen if it is in the third stadium or the last stadium, where the cancer has spread to surrounding organs. Early diagnosis of colorectal cancer is highly needed because treatment choices are decided and the period of survival is heavily affected. This paper proposes the Convolutional Neural Network (CNN) for detecting four classes of colon cancer. The data-set consists of 2500 images, divided into Tumor, Complex, Lymphoma and Stroma. This data set represents a selection from the Institute of Pathology, University of Heidelberg, Germany, consist of 150 × 150 px textures in histological pictures. The proposed system consist of two hidden convolutional layers, a fully connected layer and use Adam Optimizer with learning 0.001, and trained 10 times (epochs = 10). The result of the proposed system is 83% accuracy.

Nur Ibrahim, Nor Kumalasari Caecar Pratiwi, Muhammad Adnan Pramudito, Fauzi Frahma Taliningsih

Performance Comparison of Classification Algorithms for Locating the Dominant Heel Pain Using Electromyography Signal

The prevalence of heel pain is commonly associated with injuries, infections, nervous system problem or a non-neutral type of the foot posture. The diagnosis of heel pain can be done by a physical examination, medical records, imaging (X-Ray), Magnetic Resonance Imaging (MRI), and analysis of electromyography (EMG). An accurate diagnosis must be made to specify the appropriate treatment but it is often very difficult due to the complex anatomy of heel. The aim of this study, therefore, is to predict and compare the location of heel pain using several classification algorithms and feature extractions. In this work, the EMG signal recording is performed on three locations of human heel, namely plantar, midfoot, and posterior heel. For feature extraction step, the EMG signal is processed in the frequency domain after its frequency has been sampled (625 Hz) and filtered (butterworth filter order of 4, 10, and 300 Hz). The used extraction features are MNF, MDF, Spectral Movement, and Power Spectrum Deformation while the used classification algorithms are Support Vector Machine (SVM), adaboost, and xgboost. The results in this research are that the comparisons of prediction accuracy on heel pain location using adaboost and xgboost methods are 91 and 82%, respectively and using the linear Kernel SVM, RBF, polynomial, and sigmoid methods are 64, 73, 73, and 55%, repectively. Adaboost is the best classification algorithm among the others for predicting the location of heel pain. The developed model provides an alternative identifying tool because knowing the specific anatomic location of heel pain may help the medical practitioners in guiding the diagnosis and determining the initial or specific treatment.

Ghifari Indra Gunawan, Desri Kristina Silalahi, Husneni Mukhtar, Dandi Trianta Barus, Dien Rahmawati

The Optimal Scan Delay of Contrast Media Injection for Diagnosing Abdominal Tumors (Image Quality and Radiation Dose Aspects of Abdominal CT Scan)

Abdominal tumors due to malignancy in the liver often occurred in Abdomen CT Scan patients. To obtain a good and accurate image quality can be done by setting the scan delay and improper timing of contrast media injection. It will have an impact on the image of the tumor and also the patient dose. The purpose of this study is to determine the values of scan delay that also fulfilled the radiation protection aspects. The research was done by using 40 image samples resulted from 5 scan delay (60, 65, 70, 75, 80 s). Assessment was determined by calculated the Region of Interest (ROI) of object and dose which analyzed by Friedman tests. There was a difference in image quality. The results of dose calculations showed an increase in dose for each additional scan delay. Suggest for getting the optimal image quality with 70 s scan delay and the maximum scan delay with 75 s is still accepted for patients. The radiation dose is 24.54 mGy with maximum scan delay, which under the regulation of Bapeten and safe for the patients. The results of this study can give the recommendation for scan delay setting in abdominal CT scan to diagnose the abdominal tumor.

Siti Masrochah, Ardi Soesilo Wibowo, Jeffri Ardiyanto, Fatimah, Agung Nugroho Setiawan

Effect of mAs on the Radiation Doses Received by Eyes Organ at Cranium Examination

The study aims to see the effect of mAs to radiation doses on eye organs with the phantom cranium and to know the right mAs at cranium examination with phantom objects. So that the radiation dose received is the smallest organ. This type of research is quantitative with an experimental approach. The object was exposed using mAs; 8, 12, 16, 32, 63. Radiation dose measurements were carried out using a radiation device. The results of the first measurement that the radiation dose received by the eye organ in the phantom cranium is 0.05 mSv, the second measurement the radiation dose received by the eye organ is 0.11 mSv, the third is 0.17 mSv, the fourth measurement is 0.41 mSv, the fifth measurement is 0.77 mSv. This is influenced by mAs showing a small quantity of X-ray mAs, which will cause the number of electron clouds produced small so that when making images on the results of the radiograph, determine the amount of X-ray quantity that affects the radiation dose received by the object. A good density value is at mAs 32, producing a usable density in the normal range of 1, 23 with the radiation dose received on eyes organ are 0.41 mSv.

Ayu Wita Sari, Stanislaus Yudianus Sam

Development of Phantom Radiology Using Eggshells Powder as Bone Genu Material

Eggshells have never been used to create phantoms as bone substitutes so far. Eggshells are easy to find and commonly thrown into the waste bin. The phantom is very important in the field of radiological engineering, especially for practicums on campus to avoid X-ray radiation. The price of a phantom sold in the market is hundreds of millions. This study aims to make a Phantom using eggshell powder as a basic material for bones, especially genu, and to calculate the optical density values in bones made from eggshell powder. The main advantage of this research is that it uses a simple method and easy to find the basic materials so that the manufacturing costs are lower. By using eggshell powder, it is expected to produce Phantom products that are comparable in structure to those made in factories. The method used in this research is an experiment. Provide a comparison between the main material and the companion material; they are 1:1, 1:2, and 1:3. After the experiment is carried out, it is printed based on the appropriate comparison. The result obtained was the 1:1 ratio of the ingredients to be stronger and dry faster than the ratio of 1:2 and 1:3. The optical density value of the radiograph obtained was 0.2. Based on the references value, 0.2 on the bone radiograph included in the bright or radiopaque interval. The radiographs obtained did not show the trabecular structure. So that for the further research it is necessary to add additional materials to replace trabecular.

Ayu Wita Sari, Putri Winda Loja Bimantari, Nadela Putri Sakhia

Comparison of Radiation Dose and Image Noise in Head Computed Tomography with Sequence and Spiral Techniques

Head CT Scan techniques in clinical can use sequence and spiral technique, but the problems is dose not been routinely recorded and noise have not been assessed. This study was to determine the profile value of radiation doses and image noise in sequence and spiral technique, and also relationship between SAFIRE-Head MSCT. Quasi-experimental research design with one group pretest-posttest method to determine the effect of reducing tube voltage (kVp) 130, 110, 80 to radiation doses and image noise within and without SAFIRE. Measurements of CTDIvol, TLD and image noise. Data was analysis by Anova test followed by post hoc test. The result, there is a difference dose and noise of the spiral and sequence techniques on all variations of tube voltage. Radiation dose increases according to tube voltage and the sequence technique gives a higher dose by CTDIvol. On TLD’s measurements, doses in the eyes and thyroid are lower when using spiral techniques. High value of tube voltage produced smaller image noise. The higher level of strange (S5) SAFIRE produced lower image noise. Examination with low tube voltage in spiral technique would be reduce radiation dose (CTDIvol). However, sequence technique reduce radiation dose at the eyes and thyroid gland.

Yeti Kartikasari, Darmini, Siti Masrochah, Dwi Rochmayanti

Wrapper Subset Feature Selection for Optimal Feature Selection in Epileptic Seizure Signal Classification

Epilepsy is diagnosed by assessing the brain signal using an electroencephalograph (EEG). The assessment relies on manual visual inspection, which required experience and years of training. A computer-aided diagnose system can help neurologists assess the EEG signal. This study explores the epileptic condition by decomposing EEG signals using three levels of wavelet packet decomposition (WPD). Three orders of Daubechies mother wavelets are used. Since EEG is a non-stationary biological signal, an entropy measurement using the Shannon entropy is used to extract the signals’ information. The next process is combining the features from all levels of the decomposed signals producing 14 number of features. This study reduces the number of features using the wrapper feature subset selection (WFSS) method. The searching algorithm used is the sequential backward (SBS) and forward (SFS) selection method. The multilayer perceptron neural network (MLPNN) is used for the classification method. The system achieves the highest accuracy of 91% by using seven number of features obtained from WPD(db2) + WFSS(SBS) + MLPNN. The minimum number of features is obtained using WPD(db16) + WFSS(SFS) + MLPNN, which produces six features. While the use of WFSS(SFS) in db16 produces six features with the highest increase of accuracy by 22%. This indicates that the use of WFSS can obtain an optimal number of features set and can improve the system’s performance.

Inung Wijayanto, Rudy Hartanto, Hanung Adi Nugroho

Preliminary Study of EEG Characterization Using Power Spectral Analysis in Post-stroke Patients with Cognitive Impairment

Post-stroke dementia (PSD) or post-stroke cognitive impairment can occur in one-third of stroke sufferers. Therefore, we need a detection protocol so that patients get treatment early. Electroencephalogram (EEG) analysis is one of the methods to study deteriorating brain function where visual observation is commonly used. However, this method requires expert experience and time-consuming. Therefore, in this study, a method for characterizing EEG waves in post-stroke patients with cognitive impairment is proposed by calculating and analyzing quantitative EEG (QEEG) parameters. This study proposes a linear QEEG method through the power spectral analysis approach to characterize post-stroke patients with cognitive impairments and normal subjects. This study used a resting-awake EEG dataset collected from nineteen participants consisting of ten normal subjects, five post-stroke patients with mild cognitive impairment, and four post-stroke patients with dementia. The experiment results showed significant differences in the relative power between the three groups. These include (1) increase in delta activity and simultaneously a decrease in alpha, beta and gamma activity in dementia patients, (2) Significant differences (p-value < 0.05) on these bands are most commonly found on the frontal area electrodes and (3) there is linearity between power spectral density and the severity of dementia. This preliminary study showed that relative power analysis could be a discriminant feature among normal, post-stroke patients with mild cognitive impairment and post-stroke patients with dementia. It is hoped that the proposed method can be used to assist doctors in the early detection of post-stroke dementia and monitor the progress of dementia.

Sugondo Hadiyoso, Hasballah Zakaria, Tati Latifah E. R. Mengko, Paulus Anam Ong

Comparative Analysis of the Phonocardiogram Denoising System Based-on Empirical Mode Decomposition (EMD) and Double-Density Discrete Wavelet Transform (DDDWT)

Phonocardiogram (PCG), one of the auscultation-based technique used as a diagnostic method of the heart condition, is a patient’s heart sound recording. The simplicity, non-invasive, and passive brings an advantage to implement this method as a diagnosis system. Nevertheless, PCG recordings are often interrupted by various sources, for instance, noise from the surrounding environment, respiratory or lung sounds, power disturbances, and movement of the surrounding skin, so inhibit the PCG implementation as a diagnosis method. Therefore, it requires an appropriate method to eliminate the noise that exists in the PCG signals. To get an appropriate method in the PCG system, we compare the Empirical Mode Decomposition (EMD) and Double-Density Discrete Wavelet Transform (DD-DWT) method as a denoising system to minimize the noise effect in the PCG signal. Observation of the system performance used thirty data from the normal heart sound added by the additive white Gaussian noise (AWGN), and the performance parameter used signal-to-noise ratio (SNR) and mean square error (MSE). Based on the result, we obtained the best SNR value of 25.55 dB for the EMD method and SNR value of 18.19 dB for DD-DWT. Also, we perceived the best MSE value of 0.01% for the EMD method, and 0.42% for the DD-DWT. The results obtained show that the denoising process using the EMD method is better than the DD-DWT to implement in the PCG signal.

T. Y. Fatmawati, A. Yuliani, M. A. Afandi, D. Zulherman

Time Domain Features for EEG Signal Classification of Four Class Motor Imagery Using Artificial Neural Network

Brain-Computer Interface (BCI) is a system that measures and processes the activity of the human brain to improve or replace the function of the human body. In the future, this system can be a solution for people with disabilities, especially in locomotor organs such as hands and feet. The purpose of this research is to classify four classes of electroencephalogram (EEG) signals that represent four human motor imagery. The four motor imageries are left-hand, right-hand, left-foot, and right-foot that originated from motor imagery dataset. The proposed method in this research consists of filtering, feature extraction, and classification. The proposed method employed the Finite Impulse Response (FIR) in the filtering process to pass the required EEG signals such as delta, theta, alpha, beta, and gamma channels. The features are the Root Mean Square (RMS) values from the time domain filtered signal. Our system design used these features as input classification method that used the Artificial Neural Network (ANN). The training and testing data separation used 10-fold cross-validation. To analyze the testing performance used a confusion matrix. Based on the results, the proposed method brings the highest system accuracy as 61.2% on the beta channel.

Rahmat Widadi, Dodi Zulherman, S. Rama Febriyan Ari

Simple Visible Light Spectrophotometer Design Using 620 Nm Optical Filter

The main problem of conventional spectrophotometer is its large form and complex system. This study aims to design and develop a compact and portable visible light spectrophotometer. The main contribution of this study is to simplicity the spectrophotometer design. The designed device uses a halogen lamp as the light source, and a colored lens 620 nm optical glass filters that serve to narrow the wavelength becomes monochromatic. A phototransistor sensor is utilized to convert the monochromatic light into a voltage. A differential filter is employed to reduce 15.9 Hz noise that occurred when the phototransistor detected the light. Furthermore, all processes in the developed system are automatically controlled using ATMega328. The verification of the designed spectrophotometer is carried out using cholesterol serum with various concentrations. The conducted test result shows that the designed system has more than 90% accuracy, with the highest error percentage is 5.52%, and the mean of error percentage is 2.8%. The developed spectrophotometer demonstrated high performance, portable, and cost-effective in analyzing cholesterol serum concentrations. The proposed design of the spectrophotometer was useful in the development of simple and compact devices for chemical sample analysis.

Meilia Safitri, Farahdiba Rahmadani, Erika Loniza, Sotya Anggoro

Plasmodium Parasite Detection Using Combination of Image Processing and Deep Learning Approach

The development of an intelligent system for automated malaria detection became the one of challenges since its application supported the examination process which was conducted manually by the doctor or medical personnel. Some previous studies have been done to overcome those problems. However, most of them still have problem in detecting parasite candidates. Hence, their proposed methods did not successfully detect all parasite candidates and remains a large number of false-negative. Actually, the misdetection problem occurred since the characteristic of parasites seems unclear. To overcome these problems, we applied image processing technique and deep learning architecture to detect and to ensure whether the detected candidate is a parasite or not. Our proposed method was applied to 46 digital microscopic images provided by the Department of Parasitology, Universitas Gadjah Mada and Eijkman Institute for Molecular Biology. The proposed method comprised of four steps which are normalization process using GGB (green, green, blue) color transformation, segmentation process using Otsu followed by some morphological operations, object labelling using BLOB analysis, and classification using deep learning. Our detection process successfully detected all parasites and the classification process achieved an accuracy, sensitivity, specificity, PPV and NPV of 98.97, 100, 98.08, 97.85, and 100% respectively. This result shows that our proposed method achieved outstanding performance in both detection and classification process which indicates that our proposed method had the potential to be implemented as an intelligent system for supporting the parasitologist in conducting rapid assessment of plasmodium parasite infection.

Alifia Revan Prananda, Hanung Adi Nugroho, Eka Legya Frannita

Rapid Assessment of Breast Cancer Malignancy Using Deep Neural Network

Breast cancer is one of the deadly diseases that have high morbidity and mortality rate. Traditionally, doctors or radiologists should delineate the malignancy suspicious in a manual procedure such as manual segmentation or making diagnosis decisions. This medical examination may occur some problems such as time-consuming, tedious, and possible to produce subjective results. An alternative way to overcome this problem is that implementing technology to support this examination process. Hence, we propose a deep neural network model to reach the development of a rapid classification system for breast cancer. Our model was performed in the Breast Cancer Wisconsin (Diagnostic) dataset consisting of 569 instances and 30 attributes. Our proposed model was started by balancing the data in the preparation step. After conducting the preparation step, we performed a deep neural network model with three hidden layers and two dropout layers. Our experiment achieved the best performance compared to the previous study with accuracy, sensitivity, specificity, positive predictive value, and negative predictive value of 98.536, 98.466, 98.765, 99.689, 94.118, and 98.500% respectively which indicates that our proposed method has suitable performance for assessing cancerous of breast cancer.

Alifia Revan Prananda, Hanung Adi Nugroho, Eka Legya Frannita

Investigation on Natural Frequency of Different Thicknesses of Cartilage in Myringoplasty

A human middle ear system which includes tympanic membrane and ossicles (three tiny bones) has a function as a mechanical system transmitted sound and vibration from outer through to the inner ear section. Myringoplasty is an operation performed to repair a hole (perforation) in the tympanic membrane (TM). Sliced cartilage is generally used by an otolaryngologist to repair tympanic membrane perforation. The natural frequency for different thicknesses of cartilage in myringoplasty was compared to the normal tympanic membrane to find proper thickness in order to get the best hearing of the human ear system. The purpose of this work is to show the effect of different thicknesses of cartilage in myringoplasty using the Finite Element Method (FEM). The geometrical model of the human middle ear system was generated by the CAD Software using the physical properties of the human middle ear system reported by the previous researcher. Then, Hypermesh was used to create the Finite Element (FE) model. The eigen-value analysis is performed to obtain the mode shape and natural frequency of the human middle ear system with varied thicknesses of cartilage from 0.1 to 10 mm. Thus, by comparing the first natural frequencies of each cartilage thicknesses, we found that the maximum natural frequency of the human ear system when the cartilage’s thickness is 0.13 mm. This disorder is the effect of the human middle ear system’s mass and stiffness.

Hidayat Hidayat, Sudarsono Sudarsono, Sarwo Pranoto, Rozaini Othman

Detection of Diabetic Macular Edema in Optical Coherence Tomography Image Using Convolutional Neural Network

Diabetic macular edema (DME) is known as the main cause of patients with Diabetic Retinopathy loss their vision. The vision loss can be prevented if DME could be detected and diagnosed in the early stage. The purpose of this research is to detect DME from retinal-Optical Coherence Tomography (OCT) images using Convolutional Neural Network (CNN). In this research, 2 pre-trained models using Transfer Learning namely MobileNet and VGG-16 and 1 custom CNN models were used to classify the retinal-OCT images. The dataset of the retinal-OCT images used in this research has been obtained from Kaggle website. The dataset is organized into 3 folders (train, validation, and test) and contains subfolders for DME image category and normal image category. There are 37,663 retinal-OCT images used in the training dataset, 484 images in the validation dataset, and 16 images in the testing dataset. In this research, the custom 5 layer CNN model was compared with the 2 pre-trained models to estimate the performance of DME detection. The results show both the 2 pre-trained models using Transfer Learning and the custom 5 layer CNN model could detect DME from retinal-OCT images. Compared with the 2 pre-trained models, the custom 5 layer CNN model distinguishing DME images from normal images achieved the highest accuracy of 96%.

Sarwo Pranoto, H. Hidayat, S. Sudarsono, M. P. Lukman

Graphical User Interface for Heartbeat and Body Temperature Monitoring System Using Internet of Things (IOT)

Nowadays, the variation of the diseases revolution had become a major thread to the human kind population. The repercussion of this uncontrolled situation had led to the increasing trend regarding to the number of the patient those visited the hospital in order to obtain a treatment from the professional in the medical sector. As the number of the patient is keep increasing by the day, the number of hospital staff still not able to meet the requirement in providing the optimum service to all of their patients. The development of this project is to reduce the burden bear by the hospital staff especially the doctor by the implementation of the IOT technology. The purpose of this study is to develop a system based on the ESP8266 microcontroller that internally installed a WIFI shield provide internet connectivity easily for the user to measure the heart rate and the body temperature. By the help of Arduino environment platform, the IOT based project can be developed. This project will focus in capturing and measure the two parameters from the human body which are the heart rate and the body temperature. The sensor used are heart rate sensor, MAX30100 and the body temperature sensor, MLX90614. As the results, this system will process the raw data from the two sensors into a desire value and display the data on the OLED display and the application display. The implementation of IOT technology to this system will provide a continuous data streaming through website and showing a live development of patients for the user who is has the permission access by only using their own smart devices. Based on the finding, its shows that the comparison percentage deviation between heartbeat sensor and blood pressure monitor is 3.37% and the comparison percentage deviation between body temperature sensor and digital thermometer is 0.54%. The Graphical User Interface through Internet of Things (IOT) application system is proven to monitor accurately for heartbeat sensor measurement and body temperature sensor measurement.

Raja Siti Nur Adiimah Raja Aris, Amri Mohd Azhari, Lia Safiyah Syafie, Farah Hanan Azimi, Suzanna Ridzuan Aw

Investigation on the Effect of Varying Bubble Size and Location in Electrical Resistance Tomography Using Conducting Boundary Strategy

This paper demonstrates a Linear Back Projection (LBP) algorithm based reconstruction of conductivity distributions to investigate the effect of varying different size and location of the bubble phantom in a vertical metallic column. Sixteen electrodes were placed evenly along the circumference of the column. Simulation and experimental studies applying conducting boundary strategy were conducted to investigate the performance of the system. The reconstructed images of the phantom under test are presented. All LBP images obtained by simulation and experiment then be presented and analysed using the Mean Structural Similarity Index (MSSIM) and Area Error (AE) analysis. The number and spatial distribution of the bubble phantom can be clearly distinguished wherever they are located in the pipeline. Bubbles with a greater size than 6 mm and especially the one that is located near the wall boundary are much easier to detect. This research has successfully been applied to a metallic column and contribute the knowledge on the effect of varying both the size and location of the bubble in the column.

Suzanna Ridzuan Aw, Farah Hanan Azimi, Lia Safiyah Syafie, Ruzairi Abdul Rahim, Yasmin Abd Wahab, Nurain Izzati Shuhaimi, Raja Siti Nur Adiimah Raja Aris, Ida Laila Ahmad

Initial-to-Point of Motion Planning on Exoskeleton Arm for Post Stroke Rehabilitation

There were several statements of problem that initially rise and cause the invention. Currently, there are some problems that are still the main cause of adult disability in Malaysia and various countries. Although stroke can cause deficiencies in some neurological domains, the most normally affected is the motor system. Furthermore, given the central role normally played by hand movements in human existence, much focus is placed on remedial research focused on understanding and restoring the function of hand motor after stroke. In this study is to design arm exoskeleton for use in the field of hand rehabilitation based on the application of therapy and closely related movement training. Although the system already exists, past exoskeleton efforts involve large, expensive, invasive, and moored products and costs. Therefore, this study is m to build a cheap, smooth and effective exoskeleton system. To overcome this problem, the rehabilitation for human therapy will be a low-cost, ergonomic device actuated through sensors measuring the user’s motion. The study of exoskeleton arm is particularly interest to the researchers. Various methods have been proposed by past researchers to improve motion planning for the exoskeleton arm. However, some are still limited because the techniques are not appropriate for therapeutic applications. Motion planning is very important for the patients need to do movement as a therapist. The purpose of this study is to develop motion planning on exoskeleton arm for post-stroke rehabilitation. The motion planning from the initial to the target point. By using third order polynomial system, the trajectory planning is carried out. The starting point is known as the initial point and final point is known as target achieved. By taking the number of degrees of freedom and time, motion planning has been produced. The motion planning of exoskeleton arm is depending on the angle joint and the time to achieve the target movement.

W. T. Wan Faizura, M. Z. Muhammad Luqman, O. Mohd Hafiz, M. Safuan Naim, M. S. Muhamad Habiibullah

Fabrication of Porous Mg–Ca–Zn Alloy by High Energy Milling for Bone Implants

Biodegradable porous magnesium-based alloys are essential in hard tissue engineering, such as mechanical support during healing process, and disappear after completion of the healing process, avoiding the secondary surgery. The present study aims to fabricate Mg–Ca–Zn degradable metallic foam implant for domestically made bone replacement. The foam was introduced through the variation of TiH2 addition as blowing agent that will create a large volume fraction of gas-filled pores during heating. A High-Energy-Milling (HEM) was performed by a horizontal rotating cylinder ball mill with a rotation speed of 130 rpm for 48 h, with the ratio of metallic powders: balls are 3:7 followed by a compacting process. Some samples were subjected to preheating at 450 °C for 2 h before sintering process to observe its effect on pore formation of the green compact, followed by sintering at the temperature of 550 and 650 °C for 3 h. Samples sintered at 650 °C showed profuse micro-cracks, due to uncontrollable release of H2 gas in the liquid film creating pore rupture and MgO formation that detected from XRD analysis, that impedes the sintering process. While sintering at a lower temperature showed a dependence on preheating process. The preheated samples containing MgO suffered from micro-cracking, except for sample with 3 wt% of TiH2, where TiH2 can act as oxygen scavenger that hinders the formation of MgO. In contrast, all samples without preheating remained intact without obvious micro-cracks. Those samples have a comparable hardness value to that of the bone. The pores in the materials measured according to Archimedes procedure behave as gypsum. Thus concerning mechanical properties, the samples can be regarded as bone implants.

Ika Kartika, Doty Dewi Risanti, Hardhian Restu P. Laksana, Franciska Pramuji Lestari, Fendy Rokhmanto, Aprilia Erryani

Performance Comparison of Strain Sensors for Wearable Device in Respiratory Rate Monitoring

Respiratory rate (RR), a clinical sign of ventilation, should be measured and recorded periodically for patients on acute care. Among the other common vital signs, a change in RR can be used as the first sign of deterioration and can avoid the possibility of missing the trends’ identification as occurs in intermittent monitoring. Meanwhile some devices have experiences in false detection due to the different body motions. Hence, this study aims to develop a continuous RR monitoring with accuracy improvements using wearable sensors and data recording. In this RR monitoring work, we compared the performances of the flex sensor (2.2″ in size) and the force-sensing resistor (FSR402) by placing them on the abdomen of a relaxing subject using a belt for 6 h every 10 min. The monitoring data were not only displayed on a smartphone but also stored on firebase platform. The system therefore took the advantages of a web service architecture to send, retrieve, and access data from the cloud. The comparison results of the flex and the FSR402 sensors due to the average RR differences (and the mean average error) compared to the breath manual count are 0.30 bpm (0.90 bpm) and 1.45 bpm (2.25 bpm), respectively. In this study, we show a way to improve the sensor performance by automatically adjusting the sensitivity of the strain sensors for adapting the RR measurement on abdomen of the test subject. Furthermore, this study demonstrates that the system is able to measure and fast transmit the RR data as the monitoring process using the IoT Firebase platform and that the performance of the flex sensor provides the possibility of further development and improvement in obtaining much better result.

Ahmad Akbar Khatami, Husneni Mukhtar, Dien Rahmawati
Additional information