Unsupervised Machine Learning of the Combined Danish and Norwegian Knee Ligament Registers: Identification of 5 Distinct Patient Groups With Differing ACL Revision Rates

Abstract

Background:

Most clinical machine learning applications use a supervised learning approach using labeled variables. In contrast, unsupervised learning enables pattern detection without a prespecified outcome.

Purpose/Hypothesis:

The purpose of this study was to apply unsupervised learning to the combined Danish and Norwegian knee ligament register (KLR) with the goal of detecting distinct subgroups. It was hypothesized that resulting groups would have differing rates of subsequent anterior cruciate ligament reconstruction (ACLR) revision.

Study Design:

Cohort study; Level of evidence, 3.

Methods:

K-prototypes clustering was performed on the complete case KLR data. After performing the unsupervised learning analysis, the authors defined clinically relevant characteristics of each cluster using variable summaries, surgeons’ domain knowledge, and Shapley Additive exPlanations analysis.

Results:

Five clusters were identified. Cluster 1 (revision rate, 9.9%) patients were young (mean age, 22 years; SD, 6 years), received hamstring tendon (HT) autograft (91%), and had lower baseline Knee injury and Osteoarthritis Outcome Score (KOOS) Sport and Recreation (Sports) scores (mean, 25.0; SD, 15.6). Cluster 2 (revision rate, 6.9%) patients received HT autograft (89%) and had higher baseline KOOS Sports scores (mean, 67.2; SD, 16.5). Cluster 3 (revision rate, 4.7%) patients received bone–patellar tendon–bone (BPTB) or quadriceps tendon (QT) autograft (94%) and had higher baseline KOOS Sports scores (mean, 65.8; SD, 16.4). Cluster 4 (revision rate, 4.1%) patients received BPTB or QT autograft (88%) and had low baseline KOOS Sports scores (mean, 20.5; SD, 14.0). Cluster 5 (revision rate, 3.1%) patients were older (mean age, 42 years; SD, 7 years), received HT autograft (89%), and had low baseline KOOS Sports scores (mean, 23.4; SD, 17.6).

Conclusion:

Unsupervised learning identified 5 distinct KLR patient subgroups and each grouping was associated with a unique ACLR revision rate. Patients can be approximately classified into 1 of the 5 clusters based on only 3 variables: age, graft choice (HT, BPTB, or QT autograft), and preoperative KOOS Sports subscale score. If externally validated, the resulting groupings may enable quick risk stratification for future patients undergoing ACLR in the clinical setting. Patients in cluster 1 are considered high risk (9.9%), cluster 2 patients medium risk (6.9%), and patients in clusters 3 to 5 low risk (3.1%-4.7%) for revision ACLR.

Keywords

ACL revision outcome prediction machine learning artificial intelligence unsupervised learning

Get full access to this article

View all access options for this article.

References

Andernord

Desai

Björnsson

Ylander

Karlsson

Samuelsson

. Patient predictors of early revision surgery after anterior cruciate ligament reconstruction: a cohort study of 16,930 patients with 2-year follow-up. Am J Sports Med. 2015;43(1):121-127.

Angelini

Widera

Mobasheri

, et al. Osteoarthritis endotype discovery via clustering of biochemical marker data. Ann Rheum Dis. 2022;81(5):666-675.

Capogna

Mahure

Mollon

Duenes

Rokito

. Young age, female gender, Caucasian race, and workers’ compensation claim are risk factors for reoperation following arthroscopic ACL reconstruction. Knee Surg Sports Traumatol Arthrosc. 2020;28(7):2213-2223.

Chen

Guestrin

. XGBoost: a scalable tree boosting system. In: Proceedings of the 22nd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining. ACM; 2016:785-794.

Christensen

Krych

Engasser

Vanhees

Collins

Dahm

. Lateral tibial posterior slope is increased in patients with early graft failure after anterior cruciate ligament reconstruction. Am J Sports Med. 2015;43(10):2510-2514.

Devitt

Neri

Fritsch

. Combined anterolateral complex and anterior cruciate ligament injury: anatomy, biomechanics and management—state-of-the-art. J ISAKOS. 2023;8(1):37-46.

Eckhardt

Madjarova

Williams

, et al. Unsupervised machine learning methods and emerging applications in healthcare. Knee Surg Sports Traumatol Arthrosc. 2023;31(2):376-381.

Getgood

AMJ

Bryant

Litchfield

, et al. Lateral extra-articular tenodesis reduces failure of hamstring tendon autograft anterior cruciate ligament reconstruction: 2-year outcomes from the STABILITY study randomized clinical trial. Am J Sports Med. 2020;48(2):285-297.

Granan

Bahr

Steindal

Furnes

Engebretsen

. Development of a national cruciate ligament surgery registry: the Norwegian National Knee Ligament Registry. Am J Sports Med. 2008;36(2):308-315.

10.

Grassi

Macchiarola

Urrizola Barrientos

, et al. Steep posterior tibial slope, anterior tibial subluxation, deep posterior lateral femoral condyle, and meniscal deficiency are common findings in multiple anterior cruciate ligament failures: an MRI case-control study. Am J Sports Med. 2019;47(2):285-295.

11.

Hamrin Senorski

Svantesson

Baldari

, et al. Factors that affect patient reported outcome after anterior cruciate ligament reconstruction—a systematic review of the Scandinavian knee ligament registers. Br J Sports Med. 2019;53(7):410-417.

12.

Hartigan

Wong

. Algorithm AS 136: a K-means clustering algorithm. Appl Stat. 1979;28(1):100.

13.

Hastie

Tibshirani

Friedman

. The Elements of Statistical Learning. Springer New York; 2009.

14.

Hettrich

Dunn

Reinke

; MOON Group; Spindler

. The rate of subsequent surgery and predictors after anterior cruciate ligament reconstruction: two- and 6-year follow-up results from a multicenter cohort. Am J Sports Med. 2013;41(7):1534-1540.

15.

Kaarre

Zsidai

Narup

, et al. Scoping review on ACL surgery and registry data. Curr Rev Musculoskelet Med. 2022;15(5):385-393.

16.

Kaeding

Pedroza

Reinke

Huston

; MOON Consortium; Spindler KP. Risk factors and predictors of subsequent ACL injury in either knee after ACL reconstruction: prospective analysis of 2488 primary ACL reconstructions from the MOON cohort. Am J Sports Med. 2015;43(7):1583-1590.

17.

Pareek

, et al. Automated risk stratification of hip osteoarthritis development in patients with femoroacetabular impingement using an unsupervised clustering algorithm: a study from the Rochester Epidemiology Project. Orthop J Sports Med. 2021;9(11):232596712110506.

18.

Lundberg

Lee

. A unified approach to interpreting model predictions. Preprint published online November 24, 2017. arxiv. doi:10.48550/arXiv.1705.07874

19.

Dai

Shao

. Failed anterior cruciate ligament reconstruction: analysis of factors leading to instability after primary surgery. Chin Med J (Engl). 2013;126(2):280-285.

20.

Martin

Wastvedt

Pareek

, et al. Ceiling effect of the combined Norwegian and Danish Knee Ligament Registers limits anterior cruciate ligament reconstruction outcome prediction. Am J Sports Med. 2023;51(9):2324-2332.

21.

Martin

Wastvedt

Pareek

, et al. Machine learning algorithm to predict anterior cruciate ligament revision demonstrates external validity. Knee Surg Sports Traumatol Arthrosc. 2022;30(2):368-375.

22.

Martin

Wastvedt

Pareek

, et al. Predicting anterior cruciate ligament reconstruction revision: a machine learning analysis utilizing the Norwegian Knee Ligament Register. J Bone Joint Surg Am. 2022;104(2):145-153.

23.

Martin

Wastvedt

Pareek

, et al. Predicting subjective failure of ACL reconstruction: a machine learning analysis of the Norwegian Knee Ligament Register and patient reported outcomes. J ISAKOS. 2022;7(3):1-9.

24.

Mok

Fancher

Vopat

, et al. Sex-specific outcomes after anterior cruciate ligament reconstruction: a systematic review and meta-analysis. Orthop J Sports Med. 2022;10(2):232596712210768.

25.

MOON Knee Group; Spindler

Huston

Chagin

, et al. Ten-year outcomes and risk factors after anterior cruciate ligament reconstruction: a MOON longitudinal prospective cohort study. Am J Sports Med. 2018;46(4):815-825.

26.

Müllner

. fastcluster: fast hierarchical, agglomerative clustering routines for R and Python. J Stat Softw. 2013;53(9):1-18.

27.

Pang

Zhu

Tang

. Arthroscopic anterior cruciate ligament repair versus autograft anterior cruciate ligament reconstruction: a meta-analysis of comparative studies. Front Surg. 2022;9:887522.

28.

Persson

Fjeldsgaard

Gjertsen

, et al. Increased risk of revision with hamstring tendon grafts compared with patellar tendon grafts after anterior cruciate ligament reconstruction: a study of 12,643 patients from the Norwegian Cruciate Ligament Registry, 2004-2012. Am J Sports Med. 2014;42(2):285-291.

29.

Persson

Kjellsen

Fjeldsgaard

Engebretsen

Espehaug

Fevang

. Registry data highlight increased revision rates for Endobutton/Biosure HA in ACL reconstruction with hamstring tendon autograft: a nationwide cohort study from the Norwegian Knee Ligament Registry, 2004-2013. Am J Sports Med. 2015;43(9):2182-2188.

30.

Pruneski

Williams

Nwachukwu

, et al. The development and deployment of machine learning models. Knee Surg Sports Traumatol Arthrosc. 2022;30(12):3917-3923.

31.

Rahr-Wagner

Lind

. The Danish Knee Ligament Reconstruction Registry. Clin Epidemiol. 2016;8:531-535.

32.

Ramkumar

Pang

Polisetty

Helm

Karnuta

. Meaningless applications and misguided methodologies in artificial intelligence–related orthopaedic research propagates hype over hope. Arthroscopy. 2022;38(9):2761-2766.

33.

Riediger

Stride

Coke

Kurz

Duong

Ayeni

. ACL reconstruction with augmentation: a scoping review. Curr Rev Musculoskelet Med. 2019;12(2):166-172.

34.

Salmon

Heath

Akrawi

Roe

Linklater

Pinczewski

. 20-year outcomes of anterior cruciate ligament reconstruction with hamstring tendon autograft: the catastrophic effect of age and posterior tibial slope. Am J Sports Med. 2018;46(3):531-543.

35.

Svantesson

Hamrin Senorski

Baldari

, et al. Factors associated with additional anterior cruciate ligament reconstruction and register comparison: a systematic review on the Scandinavian knee ligament registers. Br J Sports Med. 2019;53(7):418-425.

36.

Szepannek

. clustMixType: user-friendly clustering of mixed-type data in R. R J. 2019;10(2):200.

37.

The female ACL: why is it more prone to injury? J Orthop. 2016;13(2):A1-A4.

38.

University of Cincinnati. K-means cluster analysis. UC Business Analytics R Programming Guide. Accessed January 8, 2024. https://uc-r.github.io/kmeans_clustering

39.

Vadhera

Knapik

Gursoy

, et al. Current concepts in anterior tibial closing wedge osteotomies for anterior cruciate ligament deficient knees. Curr Rev Musculoskelet Med. 2021;14(6):485-492.

40.

Webster

Feller

. Exploring the high reinjury rate in younger patients undergoing anterior cruciate ligament reconstruction. Am J Sports Med. 2016;44(11):2827-2832.

41.

Webster

Feller

Leigh

Richmond

. Younger patients are at increased risk for graft rupture and contralateral injury after anterior cruciate ligament reconstruction. Am J Sports Med. 2014;42(3):641-647.