Machine Learning Platform for Remote Analysis of Primary Health Care Technology to Support Ubiquitous Management in Clinical Engineering

This paper presents the process of using remote analysis of primary health care technology conditions through machine learning algorithms to assist the decision-making processes of ubiquitous management in clinical engineering. It includes data collection, analysis, and exhibition. This method was applied to dental technology in primary health care, a dental air compressor. The model was developed in the Microsoft Azure Machine Learning Studio platform testing the algorithms of neural networks, logistic regression, decision jungle, and decision forest, which, after comparison, was the most suitable algorithm for the purpose. The data was transformed to comprehend the influence of time in the read values to obtain an efficient result in the platform. The code execution and the exhibition of the classification results were made using a web-service and Python scripts. As a result, the system presents real-time notification and identification of health technology failures, supporting management solutions in clinical engineering.