Advances in Artificial Intelligence for automated knee osteoarthritis classification using the IKDC system

Area of Science:

• Orthopedics and Medical Imaging
• Artificial Intelligence in Healthcare

Background:

• Knee osteoarthritis is a leading cause of disability, particularly in older adults.
• Early and accurate diagnosis of knee osteoarthritis is crucial for effective patient management.
• Current diagnostic methods can be subjective and time-consuming.

Purpose of the Study:

• To develop and evaluate an AI-powered system for automated detection and classification of knee osteoarthritis.
• To utilize computer vision and machine learning for grading osteoarthritis severity based on the IKDC classification system.

Main Methods:

• A dataset of 1901 knee radiographs classified by the IKDC scale was used for training.
• A ConvNext convolutional neural network model was developed using LandingLens software.
• The model was validated on 380 test images for performance evaluation.

Main Results:

• The AI model achieved an overall accuracy of 95.16% in classifying knee osteoarthritis.
• Sensitivity was recorded at 95.11%, demonstrating robust detection capabilities.
• High class-specific accuracies (92.40%-98.45%) were observed for different osteoarthritis severity grades.

Conclusions:

• AI and computer vision offer an effective automated solution for knee osteoarthritis detection and classification.
• The developed system provides a precise and reliable tool for physicians, enhancing diagnostic efficiency.
• Integration into clinical practice can improve patient evaluation consistency and potentially lead to better health outcomes.