A novel method on earlier detection of bone cancer using Markov random field segmentation

Abstract

An enchondroma is a benign tumour found in the bone marrow and it accounts for up to 10-25% of people with benign tumours. This paper proposes an approach to detect enchondroma tumour through Magnetic Resonance (MR) images. The proposed method includes pre-processing technique such as Gaussian filter in order to remove noise and smooth images. Segmentation is the main technique which gives precise knowledge about the image using the Markov random filed approach. The feature extraction and naive Bayes classifier works effectively and detects the tumour cells which are affected by enchondroma. The proposed method works efficiently and also detects the presence of enchondroma in the MR images.