Fractal features classification for liver biopsy images using neural network-based classifier

This paper proposes the fractal features classification for liver biopsy images using probabilistic neural network (PNN). Fractal set has the properties of self-similarity and self-affinity. It can be used to estimate the fractal dimension (FD) from two-dimensional (2D) images, including the normal and cancerous liver tissue images. PNN is based on the probability density function (PDF) to implement the Bayes decision rules, and is used to develop a classifier for computer aided diagnosis. Two sets of liver biopsy images are analyzed including a normal image set and a cancerous image set. Experimental results show that the texture features can be well characterized and the PNN-based classifier has higher accuracy for pattern recognition.