Application of fuzzy c-means segmentation technique for tissue differentiation in MR images of a hemorrhagic glioblastoma multiforme

doi:10.1016/0730-725X(94)00093-I

Magnetic Resonance Imaging

Volume 13, Issue 2, 1995, Pages 277-290

https://doi.org/10.1016/0730-725X(94)00093-I Get rights and content

Abstract

The application of a raw data-based, operator-independent MR segmentation technique to differentiate boundaries of tumor from edema or hemorrhage is demonstrated. A case of a glioblastoma multiforme with gross and histopathologic correlation is presented. The MR image data set was segmented into tissue classes based on three different MR weighted image parameters (T₁-, proton density-, and T₂-weighted) using unsupervised fuzzy c-means (FCM) clustering algorithm technique for pattern recognition. A radiological examination of the MR images and correlation with fuzzy clustering segmentations was performed. Results were confirmed by gross and histopathology which, to the best of our knowledge, reports the first application of this demanding approach. Based on the results of neuropathologic correlation, the application of FCM MR image segmentation to several MR images of a glioblastoma multiforme represents a viable technique for displaying diagnostically relevant tissue contrast information used in 3D volume reconstruction. With this technique, it is possible to generate segmentation images that display clinically important neuroanatomic and neuropathologic tissue contrast information from raw MR image data.

References (26)

R.M. Haralick et al.
Survey: Image segmentation techniques
Comput. Vis. Graph. Image Process.
(1985)
M. Jungke et al.
Information processing in nuclear magnetic resonance imaging
Magn. Reson. Imaging
(1988)
H. Suzuki et al.
Automatic segmentation of head MRI images by knowledge guided thresholding
Comput. Med. Imaging Graph.
(1991)
L.P. Clarke et al.
MRI: Stability of three supervised segmentation techniques
Magn. Reson. Imaging
(1993)
M.W. Vannier et al.
Validation of magnetic resonance imaging (MRI) multispectral tissue classification
Comput. Med. Imaging Graph.
(1991)
J.L. Eilbert et al.
The variation in user drawn outlines on digital images: Effects on quantitative autoradiography
Comput. Med. Imaging Graph.
(1990)
J. Bezdek
S.K. Pal et al.
L.O. Hall et al.
A comparison of neural network and fuzzy clustering techniques in segmenting magnetic resonance images of the brain
IEEE Trans. Neural Net.
(1992)
J. Bezdek et al.
Recent convergence results for the fuzzy c-means clustering algorithms
J. Classification
(1988)

R.L. Cannon et al.

Efficient implementation of the fuzzy c-means clustering algorithm

IEEE Trans. Patt. Anal. Mach. Intell.

(1986)

M.W. Vannier et al.

Multispectral analysis of magnetic resonance images

Radiology

(1985)

M.M. O'Donnell et al.

Toward an automated analysis system for nuclear magnetic resonance imaging. II. Initial segmentation algorithm

Med. Phys.

(1986)

Cited by (153)

MRI Brain Tumor Segmentation and Analysis using Rough-Fuzzy C-Means and Shape Based Properties
2022, Journal of King Saud University - Computer and Information Sciences
Automated brain tumor segmentation of MR image is a very challenging task in a medical point of view. As the nature of the tumor, it can appear anywhere in the brain region with any size, shape, and contrast, that makes the segmentation process more difficult. In order to handle such issues, present work proposes an automated brain tumor segmentation method using rough-fuzzy C-means (RFCM) and shape based topological properties. In rough-fuzzy C-means, overlapping partition is efficiently handled by fuzzy membership and uncertainty in the datasets is resolved by lower and upper bound of the rough set. Fuzzy boundary and crisp lower approximation in RFCM play an effective contribution in brain tumor segmentation on MR images. Initial centroids selection is a major issue in C-means algorithms. Present work has introduced a method for initial centroids selection by which the execution time of RFCM is reduced as compared to random initial centroids. A patch based K-means method is also implemented for skull stripping as a preprocessing step. The proposed method was tested on MRI standard benchmark datasets. Experimental results show that the proposed method has achieved better performance based on statistical volume metrics than previous state-of-the-art algorithms with respect to ground truth (manual segmentation). It is also experimentally noticed that RFCM method achieves most promising results with higher accuracy than HCM (hard C-means) and FCM (fuzzy C-means).
A type-2 neutrosophic-entropy-fusion based multiple thresholding method for the brain tumor tissue structures segmentation
2021, Applied Soft Computing
Citation Excerpt :
Therefore, researchers employed the fuzzy c-means algorithm (FCM) [22], which was developed based on fuzzy set theory, in the segmentation of MRIs. FCM performed brain tumor segmentation in an unsupervised manner, where each pixel can have different degrees of membership in different classes [23]. Although FCM is a robust algorithm, it only works well when the images are not noisy.
In this paper, we introduce an extension theory of the neutrosophic set (NS) called type-2 neutrosophic set (T2NS). This new theory provides a granular representation of features and helps to model uncertainties with six different memberships very effectively. To demonstrate the real-time application of this theory, a new segmentation method for brain tumor tissue structures in magnetic resonance imaging (MRI) is presented. There are inconsistencies in the gray levels observed in MRIs due to their low illumination. The proposed theory addressed this problem by performing neutrosophication operation on gray levels based on six different membership functions called type-2 neutrosophic membership functions. During segmentation, a concept of T2NS entropy is used to quantify each gray level of MRIs. The proposed method is able to select multiple adaptive thresholds for segmentation of brain tumor tissue structures in MRIs from the locations of maximum entropy values of gray levels. Finally, an image fusion operation is performed on the segmented images with different thresholds to include all features and identify the location of brain tumor. The fusion images are compared with the segmented images obtained by five different methods, including fuzzy c-means algorithm, modified fuzzy c-means algorithm, fuzzy-K-means clustering algorithm, kernel intuitionistic fuzzy entropy c-means and neutrosophic entropy-based adaptive thresholding method. The proposed method achieves Jaccard similarity coefficients of 97.07%, 97.92% and 97.13% in the case of three different sets of MRIs, namely Set I, Set II and Set III, respectively. The proposed method exhibits correlation coefficients of 0.9638, 0.9698 and 0.9610 for the Set I, Set II and Set III of MRIs, respectively. Similarly, the proposed method shows uniformity measures of 0.9624, 0.9633 and 0.9660 for the Set I, Set II and Set III of MRIs, respectively. These three performance evaluation metrics show the effectiveness of the proposed method compared to the existing methods.
Brain tumour segmentation from MRI using superpixels based spectral clustering
2020, Journal of King Saud University - Computer and Information Sciences
Citation Excerpt :
Unsupervised methods for performing brain tumour segmentation include Graph Cut (Corso et al., 2008), Fuzzy-C-Means (FCM) (Emblem et al., 2009), Gaussian Mixture Model (GMM) (Liang et al. 2012), and K-means (Tuhin and Samir, 2012). In the paper (Phillips et al., 1995), the authors have used initially the fuzzy technique for brain tumour segmentation. Later, it was combined with the knowledge-based methods for more reliable performance (Emblem et al., 2009).
The automated brain tumour segmentation method is becoming challenging in the field of medical research as a brain tumour emerges with diverse size, shape and intensity. In this paper, spectral clustering is used for segmentation of brain tumour tissues from Magnetic resonance images (MRI) as it creates high-quality clusters. Spectral clustering suffers from dense similarity matrix construction for massive data. To overwhelm the drawback of spectral clustering, the proposed method performs the brain tumour segmentation by (i) identifying the tumorous region labelled as Region of Interest (ROI) using superpixel based spectral clustering. (ii) brain tumour tissues are then segmented by performing spectral clustering over the obtained ROI of MRI. The identification of ROI alleviates the computational burden of spectral clustering. The segmentation of ROI using spectral clustering produces high-quality clustering results for brain tumour segmentation. The observational results are taken out on BRATS 2012 dataset and evaluated using dice score, sensitivity and specificity metrics. The proposed method outperforms the other clustering methods with competitive dice score values for segmentation of edema and Tumor Core (TC) regions from MRI images.
Texture based localization of a brain tumor from MR-images by using a machine learning approach
2020, Medical Hypotheses
In this paper, a machine learning approach was used for brain tumour localization on FLAIR scans of magnetic resonance images (MRI). The multi-modal brain images dataset (BraTs 2012) was used, that is a skull stripped and co-registered. In order to remove the noise, bilateral filtering is applied and then texton-map images are created by using the Gabor filter bank. From the texton-map, the image is segmented out into superpixel and then the low-level features are extracted: the first order intensity statistical features and also calculates the histogram level of texton-map at each superpixel level. There is a significant contribution here that the low features are not too much significant for the localization of brain tumour from MR images, but we have to make them meaningful by integrating features with the texton-map images at the region level approach. Then these features which are provided later to classifier for the prediction of three classes: background, tumour and non-tumour region, and used the labels for computation of two different areas (i.e. complete tumour and non-tumour). A Leave-one-out (LOOCV) cross validation technique is applied and achieves the dice overlap score of 88% for the whole tumour area localization, which is similar to the declared score in MICCAI BraTS challenge. This brain tumour localization approach using the textonmap image based on superpixel features illustrates the equivalent performance with other contemporary techniques. Recently, medical hypothesis generation by using autonomous computer based systems in disease diagnosis have given the great contribution in medical diagnosis.
A survey of deep learning for MRI brain tumor segmentation methods: Trends, challenges, and future directions
2023, Health and Technology
Extended Fuzzy Possibilistic Segmentation
2023, 2023 International Conference on Power Energy, Environment and Intelligent Control, PEEIC 2023

View all citing articles on Scopus

^☆: Portions of this work have been presented as Scientific Exhibits at the American Society of Neuroradiology (ASNR) Annual Meetings in May 1993 and 1994.

View full text

Original contributionApplication of fuzzy c-means segmentation technique for tissue differentiation in MR images of a hemorrhagic glioblastoma multiforme☆

Abstract

Comput. Vis. Graph. Image Process.

Magn. Reson. Imaging

Comput. Med. Imaging Graph.

Magn. Reson. Imaging

Comput. Med. Imaging Graph.

Comput. Med. Imaging Graph.

A comparison of neural network and fuzzy clustering techniques in segmenting magnetic resonance images of the brain

IEEE Trans. Neural Net.

Recent convergence results for the fuzzy c-means clustering algorithms

J. Classification

Efficient implementation of the fuzzy c-means clustering algorithm

IEEE Trans. Patt. Anal. Mach. Intell.

Multispectral analysis of magnetic resonance images

Radiology

Toward an automated analysis system for nuclear magnetic resonance imaging. II. Initial segmentation algorithm

Med. Phys.

Original contribution
Application of fuzzy c-means segmentation technique for tissue differentiation in MR images of a hemorrhagic glioblastoma multiforme☆