Authors : K. SelvaBhuvaneswari and P. Geetha
Abstract: Semantic interpretation and understanding of medical images is an important goal of visual recognition and offers a large variety of possible applications. This research involves semantic segmentation for pixel-wise classification of images for tumor identification. Classification of brain MRI is a difficult task for tumor identification due to variance in features. Hence, the exact features that involve in the classification and identification of region of interest have to be identified. Statistical features using wavelet and semantic features using novel method are extracted from the input MRI image. These features are fed into the neurofuzzy classifier for normal and abnormal image identification. Further, the pathological tissue segmentation is done using semantic region growing approach and identification of tumor is done. The results of implementation shows the efficiency of semantic segmentation technique in identifying the pathological tissues accurately from the MRI images. The performance of the segmentation technique is evaluated by performance measures such as accuracy, specificity and sensitivity.
K. SelvaBhuvaneswari and P. Geetha, 2014. Semantic Classification and Region Growing of Brain MRI using Canfis Model for Tumor Identification. Asian Journal of Information Technology, 13: 274-281.