Summary

The application of Artificial intelligence (AI) and machine learning (ML) tools in total knee arthroplasty (TKA) has the potential to improve patient-centered decision-making and outcome prediction, as these approaches can generate patient-specific risk models. This work aims to evaluate the potential of AI and ML models in predicting TKA outcomes and identifying populations at risk.

Abstract

Background

Osteoarthritis is one of the most common causes of knee diseases, resulting in pain, reduced motion, and worse quality of life. Total Knee Arthroplasty (TKA) is one of the most well-established treatments for end-stage osteoarthritis, with an increasing number of surgeries performed worldwide. Although several advanced surgical techniques have been developed, the risk of dissatisfaction still remains high. Revision remains a major post-operative drawback. Aseptic loosening can occur in 15-20% of patients.
In orthopaedics, many studies focused on artificial intelligence (AI) and machine learning (ML) applications to optimize classification and diagnosis. However, in the field of TKA, many challenges are open to meet the clinical needs concretely. An integrated and multidisciplinary approach ranging from inflammatory to genetic and microbiome analyses and multifactorial data (clinical, psychological, demographic, kinematic, and structural) is still lacking in the management of TKA.

AIMS
This work aims to review the potential of AI and ML models in predicting TKA outcomes and identifying populations at risk.

Methods

An extensive search in the following databases: MEDLINE, Scopus, Cinahl, Google Scholar, and EMBASE was conducted using the PIOS approach to formulate the research question. The PRISMA guideline was used to report the evidence of the data extracted. A modified eight-item MINORS checklist was employed for the quality assessment. The databases were screened from the inception to June 2022.

Results

Forty-four out of the 542 initially selected articles were eligible for the data analysis; 5 further articles were identified and added to the review from the PUBMED database, for a total of 49 articles included. A total of 2,595,780 patients were identified, with an overall average age of the patients of 70.2 years ± 7.9 years old. The five most common AI/ML models identified in the selected articles were Random Forest in 38.77% of studies, Gradient Boosting Machine in 36.73% of studies, Artificial Neural Network in 34.7% of articles, Logistic Regression in 32.65%, Support Vector Machine in 26.53% of articles.

Discussion

This systematic review evaluated the possible uses of AI/ML models in TKA, highlighting their potential in improving decision-making, component sizing, inpatient costs, perioperative planning, and streamlining the surgical workflow. Implementing these prediction models in TKA can ultimately lead to more accurate predictions, less time-consuming data processing, and higher precision in identifying patterns while minimizing user input bias to provide risk-based patient-specific care. Applying an AI- or ML-based approach to a multifactorial dataset could translate multifactorial data by offering a powerful means for the identification of structural changes as well as clinical symptoms, providing a concrete tool for the personalized care of patients needing TKA.

Funded by the European Union - Next Generation EU - NRRP M6C2 - Investment 2.1 Enhancement and strengthening of biomedical research in the NHS (Project no. PNRR-MCNT2-2023-12378237- CUP: F87G24000130006)