Automatic Assessment of Lower Extremity Deformity Based on Patient Specific Computer Models

Abstract

Lower extremity deformity can cause joint pain and malfunction. Patients with severe deformity usually need a correction surgery or total knee replacement (TKR) surgery to realign the orientation of the femur and tibia. Lower extremity deformity needs to be accurately assessed before any clinical decision can be made. In practice, physicians and radiologists rely mainly on X-ray images to evaluate the deformity, and CT is used in complex cases only. Manual assessment on X-rays is tedious, time consuming and inaccurate [1]. Computer aided diagnosis was proved efficient to understand patients’ anatomy, analyze lower limb deformity and plan the possible surgery [2]. However, notable interactive works were required during the identification of the anatomic features, and the accuracy was unguaranteed when the physician is not familiar with the diagnosis software. In this study we developed an automatic assessment system to identify patients’ anatomic features and quantify lower extremity deformity.