Extended hidden Markov model for optimized segmentation of breast thermography images

Breast cancer is the most commonly diagnosed form of cancer in women. Thermography has been shown to provide an efficient screening modality for detecting breast cancer as it is able to detect small tumors and hence can lead to earlier diagnosis. This paper presents a novel extended hidden Markov model (EHMM), for optimized segmentation of breast thermogram for more effective image interpretation and easier analysis of Infrared (IR) thermal patterns. Competitive advantage of EHMM method refers to handling random sampling of the breast IR images with re-estimation of the model parameters. The performance of the algorithm is illustrated by applying EHMM segmentation method on the images of IUT_OPTIC database and compared with previously related methods. Simulation results indicate the remarkable capabilities of the proposed approach. It is worth noting that the presented algorithm is able to map semi hot regions into distinct areas and extract the regions of breast thermal images significantly, while the execution time is reduced.