Fig. 3
From: MRI grading of lumbar disc herniation based on AFFM-YOLOv8 system
AFFM-yolov8 network model, used to detect the LDH in the axis. Conv, Contains convolution layers (Conv), batch normalization (BN), and activation functions (such as SiLU), for feature extraction. The C2f (CSP with 2 convolutions) module, which replaces the C3 module of YOLOv5, further improves the efficiency of feature reuse; Spatial Pyramid Pooling Fast (SPPF): Fast spatial pyramid pooling integrates context information of different scales to enhance the robustness of the model to the target scale. Adaptive Feature Fusion Module (AFFM): improves feature fusion strategies, which can dynamically adjust fusion weights according to the importance of different feature layers.
