Detection of Deep Vein Thrombosis Using Machine Learning for CT Images
摘要
This paper’s primary goal is to provide an overview of Deep Vein Thrombosis (DVT) and its contributing factors. There is currently no clinical diagnostic for the DVT, despite it being the largest cause of disease and mortality in the globe. DVT refers to the development of blood clots in deep veins. Afflicted deep leg veins are frequently the femoral vein, popliteal vein, calf veins, or pelvic deep veins. If left untreated, it’s a potentially dangerous illness that could result in morbidity and even death. D dimer, a common laboratory screening assay, has high sensitivities but low specificities, making clinical identification of DVT difficult. More precise and noninvasive diagnostic methods are needed in order to effectively localize, monitor, and diagnose DVT in patients during and after therapy. Imaging methods are often utilized in conventional clinical therapy because they may quickly discover and identify the DVT. Due to its non-invasive nature, impedance plethysmography technology could be utilized to produce a portable DVT diagnosis equipment. On the other hand, the IP signal is susceptible to motion artifacts. Therefore, it is crucial to create MA reduction methods that are tailored to the IP signal. It can also get rid of motion with a large amplitude.