Medical imaging is currently used for diagnosis and prognosis of many diseases. Through these medical images, various phenomena that occur in the human body can be detected. In order to detect signs of a disease and various physiological activities of the human body from medical images, it is important to extract appropriate features and to interpret it properly. On the other hand, monitoring the blood flow of tissue is essential in that it can prevent the subject's prognosis and sudden death due to heart attack. Failure to monitor status of a patient can result in necrosis of the tissue, which can have a significant impact on the patient's subsequent life. Therefore, by measuring the blood flow of the tissue and understanding the blood flow characteristics of the peripheral vascular disease, it is possible to prevent the deterioration of the patients’ status. For this purpose, this study aims to extract the quantifiable features from medical images and to suggest blood flow characteristics of several peripheral vascular diseases, which also can be used as a diagnosis marker. For extracting proper features, the characteristics of each disease were considered and known imaging parameters were used. In this study, blood flow characteristics of various peripheral vascular diseases were identified and were analyzed by using known indicators, to diagnose and to state disease using medical image analysis. The practical usability has been proven through clinical practice.
After general introduction about this study (chapter 1), I proposed several feature extraction method and explained its biological meanings in chapter 2 with animal models. In chapter 3, I verified these feature extraction method is clinically useful and practically usable. Every parameters that I suggested could explain the characteristics of unknown peripheral vascular flow and had potential to use as a biomarker of each disease.
In summary, in this study, the blood flow characteristics of various peripheral vascular diseases were identified and analyzed using known indicators to diagnose and state disease with medical images, and its practical usability has been proven through clinical practice.