# An improved YOLOv10-based framework for knee MRI lesion detection with enhanced small object recognition and low contrast feature extraction

## TL;DR

This paper introduces an improved YOLOv10-based framework for detecting ACL lesions in knee MRI scans, enhancing detection of small and low-contrast lesions.

## Contribution

The novel framework integrates a C2f-SimAM module, an Adaptive Spatial Fusion module, and a hybrid loss function to improve small object recognition and low-contrast feature extraction.

## Key findings

- The model achieves a 1.3% increase in mAP@0.5 compared to standard YOLOv10 (p < 0.05).
- It improves mAP@0.5:0.95 by 2.5% over baseline models.
- The framework reduces false negatives in detecting small, low-contrast ACL tears.

## Abstract

To address the challenges in detecting anterior cruciate ligament (ACL) lesions in knee MRI examinations, including difficulties in identifying tiny lesions, insufficient extraction of low-contrast features, and poor modeling of irregular lesion morphologies, and to provide a precise and efficient auxiliary diagnostic tool for clinical practice.

An enhanced framework based on YOLOv10 is constructed. The backbone network is optimized using the C2f-SimAM module to enhance multi-scale feature extraction and spatial attention; an Adaptive Spatial Fusion (ASF) module is introduced in the neck to better fuse multi-scale spatial features; and a novel hybrid loss function combining Focal-EIoU and KPT Loss is employed. To ensure rigorous statistical evaluation, we utilized a five-fold cross-validation strategy on a dataset of 917 cases.

Evaluation on the KneeMRI dataset demonstrates that the proposed model achieves statistically significant improvements over standard YOLOv10, Faster R-CNN, and Transformer-based detectors (RT-DETR). Specifically, mAP@0.5 is increased by 1.3% (p < 0.05) compared to the standard YOLOv10, and mAP@0.5:0.95 is improved by 2.5%. Qualitative analysis further confirms the model's ability to reduce false negatives in small, low-contrast tears.

This framework effectively connects general object detection models with the specific requirements of medical imaging, providing a precise and efficient solution for diagnosing ACL injuries in routine clinical workflows.