Computer aided bone fracture detection

Abstract

Undetectable fractures and misdiagnosis are the most important problems in orthopaedics field. In recent years, researchers have studied on enhancing of diagnosis success with "Computer Aided Diagnosis" systems. In this study, "Artificial neural network (ANN) based automatic bone fracture detection system" has been performed taking into consideration mentioned needs. In this proposed system, firstly, x-ray images are taken. Secondly, these images are enhanced with the preprocessing techniques. Lastly, the new ANN based system are trained by enhanced x-ray images. Performance of this system have been evaluated over two parameters called "True Detection Rate" and "False Detection Rate". According to experimental results, the proposed system is able to detect bone fractures with 89% success rate. © 2012 IEEE.