High-uniformity strain sensor via droplet morphology engineering for anterior cruciate ligament diagnosis

The pivot shift test, widely regarded as a feasible method to assess knee joint rotational stability after anterior cruciate ligament (ACL) injury, is crucial for guiding clinical treatment planning. However, the accuracy of pivot shift test heavily relies on the operator's experience and subjective judgment due to limitations in current diagnostic tools. Here, we introduce a flexible strain sensor, the pivot shift test graphene strain sensor (PSTGSS), for ACL rotational stability assessment. For clinical hygiene and consistency, the sensor is fabricated on disposable medical polyurethane tape. By utilizing surface wettability contrast to control droplet morphology during fabrication, the strain sensors achieve high uniformity across batches. The devices demonstrate remarkable uniformity, with low relative standard deviations (RSD) in resistance (13.7 % across 25 devices), gauge factor (14.23 % across 10 devices), and bending sensitivity (18.47 % across 10 devices). In clinical applications, by analyzing the signals collected by PSTGSS, pivot shift grades (0, I, II) can be effectively distinguished, with the lag ratio (LR) strongly correlating with these grades. Furthermore, a ResNet18-based classification was performed and further refined through Grad-CAM++ interpretability analysis, resulting in a final optimized accuracy of 90.24 %. This work highlights the potential of the PSTGSS for enhancing ACL injury diagnosis accuracy via LR analysis and machine learning, and demonstrates the promising future of such strain sensors for objective evaluation in medical diagnosis.