Accuracy and longitudinal reproducibility of quantitative femorotibial cartilage measures derived from automated U-Net-b
- PDF / 4,247,857 Bytes
- 18 Pages / 595.276 x 790.866 pts Page_size
- 18 Downloads / 137 Views
RESEARCH ARTICLE
Accuracy and longitudinal reproducibility of quantitative femorotibial cartilage measures derived from automated U‑Net‑based segmentation of two different MRI contrasts: data from the osteoarthritis initiative healthy reference cohort Wolfgang Wirth1,2,3 · Felix Eckstein1,2,3 · Jana Kemnitz1 · Christian Frederik Baumgartner4 · Ender Konukoglu4 · David Fuerst1,2,3 · Akshay Sanjay Chaudhari5 Received: 9 June 2020 / Revised: 22 August 2020 / Accepted: 10 September 2020 © The Author(s) 2020
Abstract Objective To evaluate the agreement, accuracy, and longitudinal reproducibility of quantitative cartilage morphometry from 2D U-Net-based automated segmentations for 3T coronal fast low angle shot (corFLASH) and sagittal double echo at steady-state (sagDESS) MRI. Methods 2D U-Nets were trained using manual, quality-controlled femorotibial cartilage segmentations available for 92 Osteoarthritis Initiative healthy reference cohort participants from both corFLASH and sagDESS (n = 50/21/21 training/ validation/test-set). Cartilage morphometry was computed from automated and manual segmentations for knees from the test-set. Agreement and accuracy were evaluated from baseline visits (dice similarity coefficient: DSC, correlation analysis, systematic offset). The longitudinal reproducibility was assessed from year-1 and -2 follow-up visits (root-mean-squared coefficient of variation, RMSCV%). Results Automated segmentations showed high agreement (DSC 0.89–0.92) and high correlations (r ≥ 0.92) with manual ground truth for both corFLASH and sagDESS and only small systematic offsets (≤ 10.1%). The automated measurements showed a similar test–retest reproducibility over 1 year (RMSCV% 1.0–4.5%) as manual measurements (RMSCV% 0.5–2.5%). Discussion The 2D U-Net-based automated segmentation method yielded high agreement compared with manual segmentation and also demonstrated high accuracy and longitudinal test–retest reproducibility for morphometric analysis of articular cartilage derived from it, using both corFLASH and sagDESS. Keywords Cartilage · Automated segmentation · Knee osteoarthritis · Magnetic resonance imaging · Convolutional neural network
Introduction * Wolfgang Wirth [email protected] 1
Department of Imaging and Functional Musculoskeletal Research, Institute of Anatomy and Cell Biology, Paracelsus Medical University Salzburg and Nuremberg, Strubergasse 21, 5020 Salzburg, Austria
2
Ludwig Boltzmann Institute for Arthritis and Rehabilitation, Paracelsus Medical University, Salzburg, Austria
3
Chondrometrics GmbH, Ainring, Germany
4
ETH, Zurich, Switzerland
5
Department of Radiology, Stanford University, Stanford, CA, USA
Osteoarthritis (OA) is a highly prevalent, chronic disease that affects more than 300 million people world-wide [1, 2]. OA patients experience pain and functional limitations, and the knee is by far the most commonly affected joint [2]. Amongst other structural pathologies of this whole-joint-disease, articular cartilage loss is a hallmark of knee OA. Wh
Data Loading...
