Classification of breast cancer based on thermal image using support vector machine

Abstract

Advancement in computer aided diagnosis system enhances the detection competency of domain expert and reduces the time in decision making. The objective of this paper is to present the effectiveness of digital infrared thermal imaging (DITI) in the diagnosis and analysis of breast cancer and to develop an efficient method for generating nonlinear heat conduction. The proposed technique is based on the following computational methods; grey level co-occurrence matrix (GLCM) for feature extraction and support vector machine (SVM) to classify the input as cancerous or non-cancerous. Nonlinear heat conduction depends on temperature of skin surface above the tumour, and the temperature is used to investigate whether the tumour is malignant or benign. The experiments carried out on 83 images consist of 34 normal and 49 abnormal (malignant and benign tumour) from a real human breast thermal image. The classification accuracy shows 97.6 % which was significantly good.