Automated Segmentation of Knee Menisci Using U-Net Deep Learning Model: Preliminary Results

Overview

Journal Maedica (Bucur)

Specialty General Medicine

Date 2025 Feb 20

PMID 39974461

Authors

Alexei Botnari

Manuella Kadar

Daniela Rodica Puia

Jenel Marian Patrascu

Jenel Marian Patrascu Jr

Affiliations

Soon will be listed here.

Abstract

Objectives: The present study describes the initial findings of the detection and segmentation of the knee meniscus in magnetic resonance imaging (MRI) scans using the U-Net deep learning model. The primary goal was to develop a model that automatically identified and segmented the meniscus from the region of interest (ROI) in knee MRI scans.

Material And Methods: The current study was conducted in two phases. Initially, a U-Net deep learning model was developed to automatically detect the meniscus using a training dataset comprising 104 knee MRI images. In the second phase, the model was fine-tuned with an additional 50 MRI scans featuring manually segmented images to segment the meniscus from the ROI accurately.

Results: After performing 14 training tests, the U-Net model achieved a detection accuracy of 0.91. The average Dice score for ROIs after training at 100 epochs was 0.7259. With training extended to 300 epochs, the Dice score improved to 0.7525. Finally, the model reached a Dice score of 0.7609 after 500 epochs.

Conclusions: The present study introduces a practical deep learning-based approach for segmenting the knee meniscus, which is validated against ground truth annotations from orthopedic surgeons. Despite challenges such as data scarcity and the need for sequence-specific optimization, our method demonstrates significant potential for advancing automated meniscus segmentation in clinical settings.

References

Swanson M, Prescott J, Best T, Powell K, Jackson R, Haq F . Semi-automated segmentation to assess the lateral meniscus in normal and osteoarthritic knees. Osteoarthritis Cartilage. 2009; 18(3):344-53. PMC: 2826568. DOI: 10.1016/j.joca.2009.10.004. View

Byra M, Wu M, Zhang X, Jang H, Ma Y, Chang E . Knee menisci segmentation and relaxometry of 3D ultrashort echo time cones MR imaging using attention U-Net with transfer learning. Magn Reson Med. 2019; 83(3):1109-1122. PMC: 6879791. DOI: 10.1002/mrm.27969. View

Young B, Ruder J, Trofa D, Fleischli J . Visualization of Concurrent Anterolateral and Anterior Cruciate Ligament Injury on Magnetic Resonance Imaging. Arthroscopy. 2020; 36(4):1086-1091. DOI: 10.1016/j.arthro.2019.09.039. View

Olmez E, Akdogan V, Korkmaz M, Er O . Automatic Segmentation of Meniscus in Multispectral MRI Using Regions with Convolutional Neural Network (R-CNN). J Digit Imaging. 2020; 33(4):916-929. PMC: 7522137. DOI: 10.1007/s10278-020-00329-x. View

Newman A, Anderson D, Daniels A, Dales M . Mechanics of the healed meniscus in a canine model. Am J Sports Med. 1989; 17(2):164-75. DOI: 10.1177/036354658901700205. View

Kessler D, MacKay J, Crowe V, Henson F, Graves M, Gilbert F . The optimisation of deep neural networks for segmenting multiple knee joint tissues from MRIs. Comput Med Imaging Graph. 2020; 86:101793. PMC: 7721597. DOI: 10.1016/j.compmedimag.2020.101793. View

Paproki A, Engstrom C, Chandra S, Neubert A, Fripp J, Crozier S . Automated segmentation and analysis of normal and osteoarthritic knee menisci from magnetic resonance images--data from the Osteoarthritis Initiative. Osteoarthritis Cartilage. 2014; 22(9):1259-70. DOI: 10.1016/j.joca.2014.06.029. View

Zhou Z, Zhao G, Kijowski R, Liu F . Deep convolutional neural network for segmentation of knee joint anatomy. Magn Reson Med. 2018; 80(6):2759-2770. PMC: 6342268. DOI: 10.1002/mrm.27229. View

Norman B, Pedoia V, Majumdar S . Use of 2D U-Net Convolutional Neural Networks for Automated Cartilage and Meniscus Segmentation of Knee MR Imaging Data to Determine Relaxometry and Morphometry. Radiology. 2018; 288(1):177-185. PMC: 6013406. DOI: 10.1148/radiol.2018172322. View

10.

Zebis M, Aagaard P, Andersen L, Holmich P, Clausen M, Brandt M . First-time anterior cruciate ligament injury in adolescent female elite athletes: a prospective cohort study to identify modifiable risk factors. Knee Surg Sports Traumatol Arthrosc. 2021; 30(4):1341-1351. PMC: 9007777. DOI: 10.1007/s00167-021-06595-8. View

11.

Zhu W, Chern K, Mow V . Anisotropic viscoelastic shear properties of bovine meniscus. Clin Orthop Relat Res. 1994; (306):34-45. View

12.

Trunz L, Morrison W . MRI of the Knee Meniscus. Magn Reson Imaging Clin N Am. 2022; 30(2):307-324. DOI: 10.1016/j.mric.2021.11.009. View

13.

Voloshin A, Wosk J . Shock absorption of meniscectomized and painful knees: a comparative in vivo study. J Biomed Eng. 1983; 5(2):157-61. DOI: 10.1016/0141-5425(83)90036-5. View

14.

Tigges S . Diagnostic Test Metrics and Receiver Operating Characteristic Curves: The Graphic Novel. J Am Coll Radiol. 2022; 19(9):1034-1041. DOI: 10.1016/j.jacr.2022.06.011. View

15.

Pandya S, Melville D . Evaluation of the knee joint with ultrasound and magnetic resonance imaging. J Ultrason. 2023; 23(95):e239-e250. PMC: 10668946. DOI: 10.15557/jou.2023.0032. View

16.

Lecouvet F, Van Haver T, Acid S, Perlepe V, Kirchgesner T, Vande Berg B . Magnetic resonance imaging (MRI) of the knee: Identification of difficult-to-diagnose meniscal lesions. Diagn Interv Imaging. 2018; 99(2):55-64. DOI: 10.1016/j.diii.2017.12.005. View

17.

Yang Y, Ma X, An H, Liu X, An N, Ao Y . Meniscus repair with simultaneous anterior cruciate ligament reconstruction: Clinical outcomes, failure rates and subsequent processing. Chin J Traumatol. 2021; 25(1):37-44. PMC: 8787233. DOI: 10.1016/j.cjtee.2021.09.005. View