Table 11 Comparative analysis of performance metrics across model configurations (ablation study).

From: 3D CSFA-UNet: a unified attention-driven deep learning framework for accurate knee MRI segmentation and osteoarthritis severity classification

Configuration

Accuracy (%)

Precision (%)

Recall (%)

F1 score (%)

Specificity (%)

Baseline 3D U-Net

97.48

97.10

97.35

97.22

97.50

CSFA-3D UNet

98.40

98.02

98.21

98.11

98.35

CSFA-3D UNet with ASPP

98.85

98.51

98.75

98.63

98.87

CSFA-3D UNet with ASPP + DSFS

99.00

98.70

98.95

98.82

99.03

CSFA-3D UNet with ASPP + DSFS + STN

99.12

98.80

99.08

99.01

99.14

Proposed Framework

99.15

98.82

99.11

99.04

99.17