Computer-Aided Diagnosis of Type 2 Diabetes Mellitus Using Thermogram of Open Mouth

Type 2 diabetes mellitus (DM) is one of the most chronic diseases worldwide. Presently, there are no specific symptoms for this disorder. Hence, an early diagnosis is very critical to prevent the long-lasting complication caused by the disease. In this paper, the efficiency of thermogram was investigated for the evaluation of this disease when compared with biochemical standard, i.e., glycated hemoglobin (HbA(1c)). The various gray-level co-occurrence matrix (GLCM) and statistical features were extracted from the thermograms of open mouth with the help of MATLAB software. The best result obtained from classification using artificial neural network had an accuracy level of 86.7%, sensitivity of 80%, and specificity of 93.3%. The computer-aided diagnosis (CAD) model was developed for the evaluation of type 2 DM from thermogram. This study proved that the thermogram of open mouth can be an efficient, faster, and powerful diagnostic tool for type 2 DM when compared with the HbA(1c).