Research on Potential Pathological Knee Diagnosis for Undergraduates by Infrared Image Analysis

Infrared imaging has been used in clinical treatment for many years, but it is not applied widely due to its non-unique standard, unsharp imaging, and difficult feature extraction, etc. A method was presented in this study to diagnose the recessive pathological knee among the undergraduates by infrared image analysis. 200 undergraduates were chosen as the experimental subjects. The data were collected by the thermal infrared imager in a quiet room with constant temperature. The infrared images were converted into gray level images, and the gray levels of pixels in the images were converted into a sequence that correspond to the number of the pixels in each gray level. The standard curve of the multi-scale entropy value for the normal knee joint was obtained by the statistics of numerous healthy knees. Discrepancies were identified between the entropy values of abnormal knee joints and the values of the health ones. According to the differences, the degree of lesion was diagnosed. If the multi-scale entropy of the pixel series sequence in a knee infrared image was lower than that of the standard entropy in the vast majority of the scales, the knee could be diagnosed with pathological change. This infrared imaging method could be utilized for general health screening and curative effect tracing.