White Paper

White Paper: Automated Computer Diagnosis Of IR Medical Imaging

Source: FLIR Systems, Inc - Research & Science

By Marcos Leal Brioschi, Jorge Eduardo Fouto Matias, Manoel Jacobsen Teixeira, and Jose Viriato Vargas

In order to improve infrared (IR) imaging diagnosis, the use of computer software for quantitative analysis of IR images has been studied by some investigators for years. Utilizing temperature alarms alone is not sufficient for accurate diagnosis; thermal patterns tools, e.g. algorithms and fractals, can be used to identify physiological abnormalities like fever and some diseases, when adjusted for ambient and inner eye reference temperatures. The authors have developed an on-line IR image processing system with specialized algorithms to identify different diseases. Using a system of IR pattern recognition, digital geometry, and signal processing, it was possible to create a diagnostic tool that can increase the accuracy of risk analysis for breast cancer, diabetic foot ulcers, fibromyalgia thermoregulatory disturbance, knee osteoarthritis, hand/wrist rheumatoid arthritis, sleep disturbance, fever, and physiologic stress parameters. All the results were achieved from a data bank of FLIR images that the authors have compiled through the last ten years of practice. From the results obtained, the quantitative diagnosis method by a computer was found to be a significant method. The overall accuracy of a computer diagnosis may vary more or less by different diseases diagnoses. The present processing system is currently being improved by the data bank.

FLIR Systems, Inc - Research & Science