Automatic Quantification of Hip Osteoarthritis from Low-Quality X-ray Images

摘要

Diagnosis of hip osteoarthritis is conventionally done through a manual measurement of the joint distance betweenthe femoral head and the acetabular cup, a difficult and often error-prone process. Recently, Chen et al.1 proposeda fully automated technique based on landmark displacement estimation from multiple image patches that isable to accurately segment bone structures around the pelvis. This technique was shown to be comparable orbetter than state-of-the-art random forest based methods. In this paper, we report on the implementation andevaluation of this method on low-resolution datasets typically available in parts of the developing world wherehigh-resolution X-ray image technology is unavailable.We employed a dataset of hip joint images collected at a local clinic and provided to us in JPG format andat 1/3 the resolution of typical DICOM X-ray images. In addition, we employed the Dice similarity coecient,average Euclidean distance between corresponding landmarks, and Hausdorff distance to better evaluate themethod relative to diagnosis of hip osteoarthritis. Our results show that the proposed method is robust withJPEG images at 1/3 the resolution of DICOM data. Additional preliminary results quantify the accuracy ofthe approach as a function of decreasing resolution. We believe these results have important signicance forapplication in clinical settings where modern X-ray equipment is not available.