Classification of IR Image Based Inflammatory Pain Diseases Using Statistical Pattern Analysis Approach

Abstract

Thermal imaging covers a vast area to detect physiological abnormalities in a human being. Inflammation is identified by redness, swelling, warmth, and pain. Infrared (IR) imaging offers a sturdy approach to identify changes in the degree of inflammation. The aim of this work is to ascertain the extent to which statistical analysis can be used in the probabilistic classification of various inflammatory pain related diseases. Digital thermography records the skin temperature distribution of the body, thus it provides some insight regarding the type of pain associated with the inflammation. On the other hand, the corresponding pixel-based intensity distribution also provides such information. In this regard, a methodology based on statistical analysis of the underlying temperature and intensity distribution of the thermograms have been proposed. The objective of the methodology is to classify the type of pain-related diseases based on some interval estimation and identify an unknown sample with 95% confidence interval. The statistical inferences are further validated with the pathological outputs to diagnose possible conclusions.

Keywords: Infrared thermal Imaging; Arthritis; Parametric Test; Statistical Analysis