Evaluation of Automated Volumetric Cartilage Quantification for Hip Preservation Surgery

摘要

Automating the process of femoroacetabular cartilage identification from magnetic resonance imaging (MRI) images has important implications to guiding clinical care by providing a temporal metric that allows for optimizing the timing for joint preservation surgery. In this paper, we evaluate a new automated cartilage segmentation method using a time trial, segmented volume comparison, overlap metrics, and Euclidean distance mapping. We report interrater overlap metrics using the true fast imaging with steady-state precession MRI sequence of 0.874, 0.546, and 0.704 for the total overlap, union overlap, and mean overlap, respectively. This method was 3.28x faster than manual segmentation. This technique provides clinicians with volumetric cartilage information that is useful for optimizing the timing for joint preservation procedures. (C) 2016 Elsevier Inc. All rights reserved.