Automatic Classification and Segmentation of Tumors from Skull Stripped Images using PNN & SFCM
Automatic classification of brain tumor is area of concern from last few decades for better perceptive analysis in accurate manner. In this paper an automatic brain tumor classification approach namely probabilistic neural network are proposed with image and data processing techniques. The conventional algorithms which are reported in the literature are not automatic in nature and mainly their processing is based on human inspection. Then after some time a new classification approaches came into existence by overcoming the disadvantages of conventional algorithms namely Operator assisted classification methods which proves impractical for huge data amounts and simultaneously it is non-reproducible. The MR brain tumor images contains the noise like content which is mainly caused by the operator performance while processing and this noise results in highly inaccurate classification analysis. For better accuracy in classification of tumor image artificial intelligent techniques like fuzzy logic and neural networks usage are encouraged these days.
