To develop a model for detecting wrist fractures that optimizes accuracy and processing speed using advanced feature extraction and fusion techniques.
Key Findings:
PEYOLO improves mean Average Precision (mAP) by 1.4% over the baseline and shows superior performance compared to models like X-YOLO and YOLOv8-AM.
Outperforms several state-of-the-art object detection models in terms of accuracy and speed.
Demonstrates high precision and inference speed suitable for clinical applications.
Interpretation:
PEYOLO effectively addresses the challenges of detecting wrist fractures, enhancing diagnostic accuracy in clinical settings through innovative feature extraction and fusion techniques.
Limitations:
Performance may vary with different imaging conditions, such as lighting and noise levels, not covered in the dataset.
Further validation needed across diverse clinical environments to ensure robustness.
Conclusion:
PEYOLO represents a significant advancement in automated wrist fracture detection, offering improved accuracy and efficiency for clinical diagnosis.
