Thermographic Imaging System for Medical Diagnosis and Educational Process
Project no. 22288 · Status: Kapanmış (Yayın bekleniyor) · 2020
Summary
<p>One of the cancers that cause the most deaths worldwide is breast cancer. Patients' chances of survival depend significantly on early diagnosis. Today, mammography stands out as the most used method for detecting breast cancer tumors. However, this method is very expensive and may be insufficient to detect small-scale tumors. Additionally, it may bring about the harmful effects of radiation. On the other hand, infrared thermography or thermal imaging is a very low-cost and harmless method with no side effects. On the other hand, infrared thermography or thermal imaging has a very low cost, is a method with harmless side effects, and is an important candidate for early diagnosis with its higher accuracy than traditional methods. The aim of the project is the development of a complementary thermographic imaging system. Data was collected in three different environments for this aim. First, images were obtained by creating breast models in a simulation environment. Secondly, images were taken with a thermal camera from breast phantoms designed to mimic the breast organ and tumor area and the phantoms designed by embedding the vascular network. Thirdly, images were obtained from the ready database. Then, features were extracted for these three image types with asymmetry and vascularization analysis and the effective ones were identified. Finally, the ResNet50 deep network model was trained and tested separately for these three image types to separate images as healthy, early period and late period. Promising results were obtained as a result of the testing process.</p>