A computer-aided algorithm for evaluating thermograms of the hand

Abstract

The incidences of musculoskeletal cumulative trauma disorders (CTDs) of the upper extremities have been increasing for the past several years. Among the most prevalent of these disorders has been carpal tunnel syndrome (CTS). Early identification of CTS symptoms can significantly decrease the chances of injuries that are costly to the individual and to industry. One CTS diagnostic method that has gained interest is thermography. However, the interpretation of thermograms has been distinctly clinical and subjective. The purpose of this research was to develop and evaluate an algorithm that could run on a personal computer at an industrial site and objectively screen and evaluate thermograms of the hands. The algorithm worked in conjunction with a gray-scale scanner. The scanner digitized the thermal information from a Polaroid ® photograph, and the algorithm constructed histograms based on the temperature distributions of the digitized thermographic data. Thermograms from the palm side of the hands of 10 female patients were made and evaluated by a physician as asymptomatic (having no symptoms) of median nerve injuries. After the algorithm calculated the temperature histograms and temperature means of the thermographic data, it then compared them statistically. Although this technique appears to hold significant promise for objectively evaluating thermograms of the hands, using color rather than gray-scale scanning technology would upgrade it substantially.