tejaswini reddy - Academia.edu (original) (raw)
Related Authors
Centre National de la Recherche Scientifique / French National Centre for Scientific Research
Uploads
Papers by tejaswini reddy
A novel approach called enhanced possibilistic fuzzy c-means clustering is proposed for segmentin... more A novel approach called enhanced possibilistic fuzzy c-means clustering is proposed for segmenting MRI brain image into different tissue types on both normal and tumor-affected pathological brain images. The proposed method incorporates membership, possibility (typicality) and both local and non-local spatial neighborhood information to classify each pixel by combining the fuzzy c-mean (FCM) and possibilistic c-mean. This incorporation is achieved by modifying the distance metrics. This improves accuracy of the medical image segmentation in both real and noisy images. Application of our method to contrast-enhanced T1-weighted brain images gives segmentation of white matter, gray matter and cerebrospinal fluid of brain image. The average value of similarity metrics of the result obtained from our method is 96 %. This value that is higher than the other methods shows that our proposed method segments the MRI brain image effectively. Experimental results with synthetic and real images show that the proposed algorithm is more accurate and robust than other FCM clustering algorithm extension.
A novel approach called enhanced possibilistic fuzzy c-means clustering is proposed for segmentin... more A novel approach called enhanced possibilistic fuzzy c-means clustering is proposed for segmenting MRI brain image into different tissue types on both normal and tumor-affected pathological brain images. The proposed method incorporates membership, possibility (typicality) and both local and non-local spatial neighborhood information to classify each pixel by combining the fuzzy c-mean (FCM) and possibilistic c-mean. This incorporation is achieved by modifying the distance metrics. This improves accuracy of the medical image segmentation in both real and noisy images. Application of our method to contrast-enhanced T1-weighted brain images gives segmentation of white matter, gray matter and cerebrospinal fluid of brain image. The average value of similarity metrics of the result obtained from our method is 96 %. This value that is higher than the other methods shows that our proposed method segments the MRI brain image effectively. Experimental results with synthetic and real images show that the proposed algorithm is more accurate and robust than other FCM clustering algorithm extension.