Related Experiment Video
Updated: Jun 12, 2025
08:09
Osteoarthritis Pain Model Induced by Intra-Articular Injection of Mono-Iodoacetate in Rats
Published on: May 20, 2020
8.3K
Investigating artificial intelligence models for predicting joint pain from serum biochemistry
Saman Shahid1, Aatir Javaid2, Usman Amjad2
1National University of Computer and Emerging Sciences, Foundation for the Advancement of Science and Technology, Department of Sciences and Humanities - Lahore, Pakistan.
Revista Da Associacao Medica Brasileira (1992)
|September 18, 2024
Summary
Machine learning models accurately predict joint pain using patient data. Uric acid levels were the strongest indicator, enabling early detection and prevention of orthopedic issues.
Area of Science:
- Orthopedics
- Artificial Intelligence
- Machine Learning
Background:
- Joint swelling and myalgia are common orthopedic complaints.
- Early detection of joint pain is crucial for preventing severe complications.
- Machine learning offers potential for accurate diagnostic tools.
Purpose of the Study:
- To evaluate machine learning models for predicting joint pain.
- To identify key clinical attributes for joint pain prediction.
- To assess the performance of different machine learning algorithms.
Main Methods:
- Patients with joint swelling/myalgia were analyzed.
- Clinical data including uric acid, CRP, and blood counts were collected.
- Random Forest, Gradient Boosted, Multilayer Perceptron, and Radial Basis Function models were employed.
Main Results:
- Random Forest achieved 97% accuracy in predicting joint pain.
- Multilayer Perceptron demonstrated 98% accuracy, outperforming Radial Basis Function.
- Uric acid was the most significant predictor (100% relevance), followed by creatinine and AST.
Conclusions:
- Artificial intelligence-based detection of joint pain is feasible.
- Early diagnosis via AI can prevent advanced orthopedic conditions.
- Machine learning models, particularly Multilayer Perceptron, show promise for clinical application.
