TY - GEN
T1 - Fuzzy C-means applied to MRI images for an automatic lesion detection using image enhancement and constrained clustering
AU - Kinani, Jean Marie Vianney
AU - Rosales-Silva, Alberto J.
AU - Gallegos-Funes, Francisco J.
AU - Arellano, Alfonso
N1 - Publisher Copyright:
© 2014 IEEE.
PY - 2015/1/6
Y1 - 2015/1/6
N2 - In this work, we present a fast and robust practical tool for detection of lesions with minimal user interaction. Particularly, a fuzzy image enhancement is performed on both T1 weighted magnetic resonance (MR), and Fluid attenuated inversion recovery (FLAIR) images to facilitate a better segmentation. We establish a fuzzy system that performs the intensity transformation through the implication method; after, the scalar output obtained from this system is used to separate healthy from the unhealthy structures using constrained fuzzy clustering. An advantage of this lesion detection pipeline is the simultaneous use of features computed from the intensity properties of the image in a cascading pattern, which makes the computation fast, robust and self-contained. We empirically validate our algorithm with large scale experiments using both clinical and synthetic brain lesion datasets, and an 84%-93% overlap performance of the proposed algorithm was attained with an emphasis on robustness with respect to different and heterogeneous lesion types, and its effectiveness in terms of computation time.
AB - In this work, we present a fast and robust practical tool for detection of lesions with minimal user interaction. Particularly, a fuzzy image enhancement is performed on both T1 weighted magnetic resonance (MR), and Fluid attenuated inversion recovery (FLAIR) images to facilitate a better segmentation. We establish a fuzzy system that performs the intensity transformation through the implication method; after, the scalar output obtained from this system is used to separate healthy from the unhealthy structures using constrained fuzzy clustering. An advantage of this lesion detection pipeline is the simultaneous use of features computed from the intensity properties of the image in a cascading pattern, which makes the computation fast, robust and self-contained. We empirically validate our algorithm with large scale experiments using both clinical and synthetic brain lesion datasets, and an 84%-93% overlap performance of the proposed algorithm was attained with an emphasis on robustness with respect to different and heterogeneous lesion types, and its effectiveness in terms of computation time.
KW - Fuzzy C-means
KW - MRI images
KW - Medical imaging
KW - Tumor detection
UR - http://www.scopus.com/inward/record.url?scp=84921716691&partnerID=8YFLogxK
U2 - 10.1109/IPTA.2014.7001987
DO - 10.1109/IPTA.2014.7001987
M3 - Contribución a la conferencia
AN - SCOPUS:84921716691
T3 - 2014 4th International Conference on Image Processing Theory, Tools and Applications, IPTA 2014
BT - 2014 4th International Conference on Image Processing Theory, Tools and Applications, IPTA 2014
A2 - Djemal, Khalifa
PB - Institute of Electrical and Electronics Engineers Inc.
T2 - 4th International Conference on Image Processing Theory, Tools and Applications, IPTA 2014
Y2 - 14 October 2014 through 17 October 2014
ER -