Thermal image analysis and segmentation of hand in evaluation of rheumatoid arthritis

Rheumatoid arthritis (RA) is a chronic inflammatory disease that affects and destroys the joints of fingers, wrist and feet. Although different imaging modalities like x-ray, magnetic resonance imaging and ultrasound are available for diagnosing the RA. Thermal imaging is considered as a novel imaging technique for diagnosing the RA. Thermal imaging technique is based on infrared Thermograms depicting the temperature variations in abnormal region of interest. The objectives of this study was i) to evaluate the rheumatoid arthritis based on heat distribution index and skin temperature measurements and to analyse the difference in skin temperature measurement in hand for RA patients and normal persons. ii) to automatically segment the abnormal regions of the hand especially for arthritis patients using fuzzy c means algorithm and Expectation Maximization (EM) algorithm. In this paper, thermal image analysis was done based on heat distribution index(HDI) and skin temperature measurement. The heat distribution value is obtained as 1.53±0.5 From the temperature analysis the results predicted was there is an increase in temperature of 0.96°c in hand region of RA patients compared to normal patients. The correlation between HDI and skin temperature measurement was statistically significant (r=0.63, p<;0.05). Fuzzy c-means algorithm has better results compared to EM Algorithm in evaluating the disease.