Damarsal dejenerasyon tespitinin artırılmış gerçeklik ortamına entegrasyonu: Yakın-kızılötesi spektroskopi ve derin öğrenme temelli bir e-sağlık uygulaması

Developing technology makes our lives easier in every area imaginable. Today, the integration of images obtained with medical imaging techniques, especially in the deep learning infrastructure, mediates the saving of countless lives within the scope of early diagnosis and follow-up. In this context, the star of e-health applications that especially appeal to home use is shining rapidly. The images obtained from the user at periodic intervals in the Imaging Technique Phase of this thesis, are enhanced on the server with the Digital Image Pre-Processing Phase. Visual deficiencies caused by visualization are corrected to a certain extent in the Digital Image Post-Processing Phase. If the Classification Phase confirms that the same tissue region is being viewed, vascular degeneration (narrowing/enlargement) in the relevant images is detected in the Object Detection Phase. Videos showing the degeneration detection results and their locations are prepared by the Augmented Reality Phase and shared with the user and the physician. In this way, possible degeneration can be detected at an early stage and its developmental stages can be followed. According to the experimental results, the most effective object detection was obtained for varicose_vein class with Accuracy Rate (1.0000), Misclassification Rate (0.0000) and Precision (1.0000). In general, within the scope of this study, molecules that can be revealed by using near-infrared radiation that cannot be detected by the eye, computer systems that can create their own intelligence with deep learning, and augmented reality environments that can offer time-independent visualization have been examined.