Segmentation of knee MRI data with convolutional neural networks for semi-automated three-dimensional surface-based analysis of cartilage morphology and composition

Kessler, Dimitri A., Mackay, James W., McDonnell, Stephen M., Janiczek, Robert L., Graves, Martin J., Kaggie, Joshua D. and Gilbert, Fiona J. (2022) Segmentation of knee MRI data with convolutional neural networks for semi-automated three-dimensional surface-based analysis of cartilage morphology and composition. Osteoarthritis Imaging, 2 (2). ISSN 2772-6541

[img]
Preview
PDF (1-s2.0-S2772654122000046-main) - Accepted Version
Available under License Creative Commons Attribution Non-commercial No Derivatives.

Download (2MB) | Preview

Abstract

Objective: To assess automatic segmentations for surface-based analysis of cartilage morphology and composition on knee magnetic resonance (MR) images. Methods: 2D and 3D U-Nets were trained on double echo steady state (DESS) images from the publicly available Osteoarthritis Initiative (OAI) dataset with femoral and tibial bone and cartilage segmentations provided by the Zuse Institute Berlin (ZIB). The U-Nets were used to perform automatic segmentation of femoral and tibial bone-cartilage structures (bone and cartilage segmentations combined into one structure) from the DESS images. T2-weighted images from the OAI dataset were registered to the DESS images and used for T2 map calculation. Using the 3D cartilage surface mapping (3D-CaSM) method, surface-based analysis of cartilage morphology (thickness) and composition (T2) was performed using both manual and network-generated segmentations from OAI ZIB testing images. Bland-Altman analyses were performed to evaluate the accuracy of the extracted cartilage thickness and T2 measurements from both U-Nets compared to manual segmentations. Results: Bland-Altman analysis showed a mean bias [95% limits of agreement] for femoral and tibial cartilage thickness measurements ranging between -0.12 to 0.33 [-0.28, 0.96] mm with 2D U-Net and 0.07 to 0.14 [-0.14, 0.39] mm with 3D U-Net. For T2, the mean bias [95% limits of agreement] ranged between -0.16 to 1.32 [-4.71, 4.83] ms with 2D U-Net and -0.05 to 0.46 [-2.47, 3.39] ms with 3D U-Net. Conclusions: While both 2D and 3D U-Nets exemplified the time-efficiency benefit of using deep learning methods for generating the required segmentations, segmentations from 3D U-Nets demonstrated higher accuracy in the extracted thickness and T2 features using 3D-CaSM compared to the segmentations from 2D U-Nets.

Item Type: Article
Additional Information: Acknowledgments: The authors acknowledge research support from GlaxoSmithKline, the National Institute of Health Research (NIHR) Cambridge Biomedical Research Centre (BRC-1215-20014) and the Addenbrooke's Charitable Trust (ACT). JDK acknowledges support from Horizon EU 2020 (grant agreement no. 761214). The views expressed in this article are those of the author(s) and not necessarily those of GlaxoSmithKline, the NIHR, the ACT or the Department of Health and Social Care. This manuscript was prepared using public use data sets from the Osteoarthritis Initiative (OAI) and Zuse Institute Berlin (ZIB) and does not necessarily reflect the opinions or views of the OAI investigators, the NIH, the private funding partners or the ZIB.
Faculty \ School: Faculty of Medicine and Health Sciences > Norwich Medical School
Related URLs:
Depositing User: LivePure Connector
Date Deposited: 29 Mar 2022 09:30
Last Modified: 24 May 2022 14:53
URI: https://ueaeprints.uea.ac.uk/id/eprint/84323
DOI: 10.1016/j.ostima.2022.100010

Actions (login required)

View Item View Item