Magnetic resonance imaging (MRI) has been proposed as a complimentary method to measure bone quality and assess fracture risk. However, manual segmentation of MR images of bone is time-consuming, limiting the use of MRI measurements in the clinical practice. The purpose of this paper is to present an automatic proximal femur segmentation method that is based on deep convolutional neural networks (CNNs). This study had institutional review board approval and written informed consent was obtained from all subjects. A dataset of volumetric structural MR images of the proximal femur from 86 subject were manually-segmented by an expert. We performed experiments by training two different CNN architectures with multiple number of initial feature maps and layers, and tested their segmentation performance against the gold standard of manual segmentations using four-fold cross-validation. Automatic segmentation of the proximal femur achieved a high dice similarity score of 0.94$\pm$0.05 with precision = 0.95$\pm$0.02, and recall = 0.94$\pm$0.08 using a CNN architecture based on 3D convolution exceeding the performance of 2D CNNs. The high segmentation accuracy provided by CNNs has the potential to help bring the use of structural MRI measurements of bone quality into clinical practice for management of osteoporosis.
Submitted 20 Apr 2017 to Computer Vision and Pattern Recognition
Published 21 Apr 2017
Updated 14 Feb 2018
Author comments: 26 pages, 5 figures, and 2 tableshttp://arxiv.org/abs/1704.06176http://arxiv.org/pdf/1704.06176.pdf