A system to help physicians with fracture detection in stents

Fractures in stents are usually detected by visual analysis, which may be affected by the presence of noise and image deformations. The lack of research into automating stent fracture detection has motivated this work, in which techniques are developed to facilitate diagnosis by observation (Image Delineation Algorithm) and, when possible, to point out areas of possible fractures (Fracture Detection Algorithm). The use of classical elements and the development of additional computational techniques contributed to the process of image analysis, providing a possible way to aid medical diagnosis. The developed algorithms are applied to image samples from femoropopliteal arteries, and the results are compared to those of medical diagnosis. As a result, aside from the improvement of image display, a kappa concordance index of 0.878 for the detection of fractures confirms the method as satisfactory, with very good agreement with medical diagnosis.