Deep learning-assisted detection of meniscus and anterior cruciate ligament combined tears in adult knee magnetic resonance imaging: a crossover study with arthroscopy correlation

被引:1
作者
Behr, Julien [1 ,2 ]
Nich, Christophe [1 ,3 ]
D'Assignies, Gaspard [4 ,5 ]
Zavastin, Catalin [6 ]
Zille, Pascal [4 ]
Herpe, Guillaume [4 ,7 ]
Triki, Ramy [1 ]
Grob, Charles [2 ]
Pujol, Nicolas [2 ]
机构
[1] Nantes Univ, CHU Nantes, Clin Chirurg Orthoped & Traumatol, Nantes, France
[2] Univ Versailles St Quentin En Yvelines, Hop Mignot, Ctr Hosp Versailles, Serv Chirurg Orthoped & Traumatol, Versailles, France
[3] Nantes Univ, INSERM, UMRS 1229, Regenerat Med & Skeleton RMeS,ONIRIS, Nantes, France
[4] Incepto Med, Paris, France
[5] Grp Hosp Havre, Serv Radiol, Le Havre, France
[6] Univ Versailles St Quentin En Yvelines, Hop Mignot, Ctr Hosp Versailles, Serv Radiol, Versailles, France
[7] LAbCom I3M DACTIM MIS, CNRS 7348, Poitiers, France
关键词
Machine learning; Deep learning; AI; MRI; Knee; Meniscus; Anterior cruciate ligament; MRI; DIAGNOSIS; RECONSTRUCTION; EXPERIENCE; ACCURACY; LESIONS; MOON;
D O I
10.1007/s00264-025-06531-2
中图分类号
R826.8 [整形外科学]; R782.2 [口腔颌面部整形外科学]; R726.2 [小儿整形外科学]; R62 [整形外科学(修复外科学)];
学科分类号
摘要
AimWe aimed to compare the diagnostic performance of physicians in the detection of arthroscopically confirmed meniscus and anterior cruciate ligament (ACL) tears on knee magnetic resonance imaging (MRI), with and without assistance from a deep learning (DL) model.MethodsWe obtained preoperative MR images from 88 knees of patients who underwent arthroscopic meniscal repair, with or without ACL reconstruction. Ninety-eight MR images of knees without signs of meniscus or ACL tears were obtained from a publicly available database after matching on age and ACL status (normal or torn), resulting in a global dataset of 186 MRI examinations. The Keros (R) (Incepto, Paris) DL algorithm, previously trained for the detection and characterization of meniscus and ACL tears, was used for MRI assessment. Magnetic resonance images were individually, and blindly annotated by three physicians and the DL algorithm. After three weeks, the three human raters repeated image assessment with model assistance, performed in a different order.ResultsThe Keros (R) algorithm achieved an area under the curve (AUC) of 0.96 (95% CI 0.93, 0.99), 0.91 (95% CI 0.85, 0.96), and 0.99 (95% CI 0.98, 0.997) in the detection of medial meniscus, lateral meniscus and ACL tears, respectively. With model assistance, physicians achieved higher sensitivity (91% vs. 83%, p = 0.04) and similar specificity (91% vs. 87%, p = 0.09) in the detection of medial meniscus tears. Regarding lateral meniscus tears, sensitivity and specificity were similar with/without model assistance. Regarding ACL tears, physicians achieved higher specificity when assisted by the algorithm (70% vs. 51%, p = 0.01) but similar sensitivity with/without model assistance (93% vs. 96%, p = 0.13).ConclusionsThe current model consistently helped physicians in the detection of medial meniscus and ACL tears, notably when they were combined.Level of evidenceDiagnostic study, Level III.
引用
收藏
页码:1689 / 1697
页数:9
相关论文
共 33 条
[1]   Deep-learning-assisted diagnosis for knee magnetic resonance imaging: Development and retrospective validation of MRNet [J].
Bien, Nicholas ;
Rajpurkar, Pranav ;
Ball, Robyn L. ;
Irvin, Jeremy ;
Park, Allison ;
Jones, Erik ;
Bereket, Michael ;
Patel, Bhavik N. ;
Yeom, Kristen W. ;
Shpanskaya, Katie ;
Halabi, Safwan ;
Zucker, Evan ;
Fanton, Gary ;
Amanatullah, Derek F. ;
Beaulieu, Christopher F. ;
Riley, Geoffrey M. ;
Stewart, Russell J. ;
Blankenberg, Francis G. ;
Larson, David B. ;
Jones, Ricky H. ;
Langlotz, Curtis P. ;
Ng, Andrew Y. ;
Lungren, Matthew P. .
PLOS MEDICINE, 2018, 15 (11)
[2]   Intra-articular Findings in Primary and Revision Anterior Cruciate Ligament Reconstruction Surgery A Comparison of the MOON and MARS Study Groups [J].
Borchers, James R. ;
Kaeding, Christopher C. ;
Pedroza, Angela D. ;
Huston, Laura J. ;
Spindler, Kurt P. ;
Wright, Rick W. .
AMERICAN JOURNAL OF SPORTS MEDICINE, 2011, 39 (09) :1889-1893
[3]   Deep Learning for Detection of Complete Anterior Cruciate Ligament Tear [J].
Chang, Peter D. ;
Wong, Tony T. ;
Rasiej, Michael J. .
JOURNAL OF DIGITAL IMAGING, 2019, 32 (06) :980-986
[4]   Magnetic resonance imaging versus arthroscopy in the diagnosis of knee pathology, concentrating on meniscal lesions and ACL tears: a systematic review [J].
Crawford, Ruth ;
Walley, Gayle ;
Bridgman, Stephen ;
Maffulli, Nicola .
BRITISH MEDICAL BULLETIN, 2007, 84 :5-23
[5]   Clinical and MRI findings associated with false-positive knee MR diagnoses of medial meniscal tears [J].
De Smet, Arthur A. ;
Nathan, David H. ;
Graf, Ben K. ;
Haaland, Benjamin A. ;
Fine, Jason P. .
AMERICAN JOURNAL OF ROENTGENOLOGY, 2008, 191 (01) :93-99
[6]   MENISCAL TEARS MISSED ON MR-IMAGING - RELATIONSHIP TO MENISCAL TEAR PATTERNS AND ANTERIOR CRUCIATE LIGAMENT TEARS [J].
DESMET, AA ;
GRAF, BK .
AMERICAN JOURNAL OF ROENTGENOLOGY, 1994, 162 (04) :905-911
[7]   The Fate of Meniscus Tears Left In Situ at the Time of Anterior Cruciate Ligament Reconstruction: A 6-Year Follow-up Study From the MOON Cohort [J].
Duchman, Kyle R. ;
Westermann, Robert W. ;
Spindler, Kurt P. ;
Reinke, Emily K. ;
Huston, Laura J. ;
Amendola, Annunziato ;
Wolf, Brian R. .
AMERICAN JOURNAL OF SPORTS MEDICINE, 2015, 43 (11) :2688-2695
[8]   Accuracy of MRI evaluation of meniscus tears in the setting of ACL injuries [J].
Dufka, Faustine L. ;
Lansdown, Drew A. ;
Zhang, Alan L. ;
Allen, Christina R. ;
Ma, C. Benjamin ;
Feeley, Brian T. .
KNEE, 2016, 23 (03) :460-464
[9]   Artificial intelligence for MRI diagnosis of joints: a scoping review of the current state-of-the-art of deep learning-based approaches [J].
Fritz, Benjamin ;
Fritz, Jan .
SKELETAL RADIOLOGY, 2022, 51 (02) :315-329
[10]  
Fritz B, 2020, SKELETAL RADIOL, V49, P1207, DOI 10.1007/s00256-020-03410-2