Harbiodun0122/Knee-Osteoarthritis-Severity-Grading
Knee Osteoarthritis Severity Grading
The project intersects artificial intelligence and health care, focusing on creating deep learning models that can detect knee joints in medical images and classify the severity of osteoarthritis in the detected joint. It leverages an object detection framework (Detectron2) for the detection of knee joints and a deep learning classification model (Xception) for osteoarthritis severity classification.
Detectron2 Performance metrics
- Average Precision (AP) @[ IoU=0.50:0.95] = 0.407
- Average Precision (AP) @[ IoU=0.50] = 0.849
- Average Recall (AR) @[ IoU=0.50:0.95] = 0.458
Xception Performance metrics
Accuracy (sklearn): 0.8870588235294118 Classification Report: precision recall f1-score support
0 0.89 0.92 0.90 85 1 0.84 0.84 0.84 85 2 0.80 0.91 0.85 85 3 0.95 0.87 0.91 85 4 0.99 0.91 0.94 85
accuracy 0.89 425 macro avg 0.89 0.89 0.89 425 weighted avg 0.89 0.89 0.89 425