Auto-segmentation of the tibia and femur from knee MR images via deep learning and its application to cartilage strain and recovery

• Published in: Journal of biomechanics Vol. 149; p. 111473
• Main Authors: Kim-Wang, Sophia Y., Bradley, Patrick X., Cutcliffe, Hattie C., Collins, Amber T., Crook, Bryan S., Paranjape, Chinmay S., Spritzer, Charles E., DeFrate, Louis E.
• Format: Journal Article
• Language: English
• Published: United States Elsevier Ltd 01.03.2023; Elsevier Limited
• Subjects: Algorithms; Artificial neural networks; Auto-segmentation; Business metrics; Cartilage; Computer architecture; Confidence intervals; Deep Learning; Femur; Femur - diagnostic imaging; Humans; Image Processing, Computer-Assisted - methods; Image segmentation; Knee; Knee Joint - diagnostic imaging; Learning algorithms; Machine Learning; Magnetic resonance imaging; Magnetic Resonance Imaging - methods; MRI; Neural networks; Recovery; Soft tissues; Statistical analysis; Strain; Three dimensional models; Tibia; Tibia - diagnostic imaging; U-Net; United States > US; Massachusetts; Auto-segmentation; MRI; Machine Learning; U-Net; Strain
• ISSN: 0021-9290; 1873-2380; 1873-2380
• DOI: 10.1016/j.jbiomech.2023.111473

The ability to efficiently and reproducibly generate subject-specific 3D models of bone and soft tissue is important to many areas of musculoskeletal research. However, methodologies requiring such models have largely been limited by lengthy manual segmentation times. Recently, machine learning, and more specifically, convolutional neural networks, have shown potential to alleviate this bottleneck in research throughput. Thus, the purpose of this work was to develop a modified version of the convolutional neural network architecture U-Net to automate segmentation of the tibia and femur from double echo steady state knee magnetic resonance (MR) images. Our model was trained on a dataset of over 4,000 MR images from 34 subjects, segmented by three experienced researchers, and reviewed by a musculoskeletal radiologist. For our validation and testing sets, we achieved dice coefficients of 0.985 and 0.984, respectively. As further testing, we applied our trained model to a prior study of tibial cartilage strain and recovery. In this analysis, across all subjects, there were no statistically significant differences in cartilage strain between the machine learning and ground truth bone models, with a mean difference of 0.2 ± 0.7 % (mean ± 95 % confidence interval). This difference is within the measurement resolution of previous cartilage strain studies from our lab using manual segmentation. In summary, we successfully trained, validated, and tested a machine learning model capable of segmenting MR images of the knee, achieving results that are comparable to trained human segmenters.