A novel image processing procedure for thermographic image analysis

The imaging procedure shown in this paper has been developed for processing thermographic images, measuring the ocular surface temperature (OST) and visualizing the ocular thermal maps in a fast, reliable, and reproducible way. The strength of this new method is that the measured OSTs do not depend on the ocular geometry; hence, it is possible to compare the ocular profiles belonging to the same subject (right and left eye) as well as to different populations. In this paper, the developed procedure is applied on two subjects’ eyes: a healthy case and another affected by an ocular malignant lesion. However, the method has already been tested on a bigger group of subjects for clinical purpose. For demonstrating the potentiality of this method, both intra- and inter-examiner repeatability were investigated in terms of coefficients of repeatability (COR). All OST indices showed repeatability with small intra-examiner (%COR 0.06–0.80) and inter-examiner variability (%COR 0.03–0.94). Measured OSTs and thermal maps clearly showed the clinical condition of the eyes investigated. The subject with no ocular pathology had no significant difference (P value = 0.25) between the OSTs of the right and left eye. On the contrary, the eye affected by a malignant lesion was significantly warmer (P value < 0.0001) than the contralateral, where the lesion was located. This new procedure demonstrated its reliability; it is featured by simplicity, immediacy, modularity, and genericity. The latter point is extremely precious as thermography has been used, in the last decades, in different clinical applications.