Abstract
The hamstring muscle group is the most frequently injured muscle group in non-contact muscle injuries in sports involving high-speed running. A total of 84% of hamstring injuries affect the biceps femoris (BF) muscle. Clinical assessments and magnetic resonance imaging (MRI) are routinely used for diagnosis and plan management. MRI-negative scans for clinically diagnosed hamstring injuries range from 14% to 45%. We tested the hypothesis that the functional differences between injured and non-injured BF assessed by tensiomyography can be used for diagnostic and classification purposes. We compared an injured group of 53 international-level soccer players and sprinters with 53 non-injured international-level soccer players and sprinters of both sexes. Comparing the injured vs. non-injured athletes and the left vs. right side in all of the athletes, we used the percentage of absolute differences in the BF contraction time (Tc) to classify non-injured and injured BF muscles. The receiver operating characteristic (ROC) curve and the area under the curve (AUC) and the precision−recall curve (PRC) were used to measure the classification accuracy and to identify cut-off limits using the Tc differences. There was a very high ROC AUC value of 0.981 (SE = 0.009, p < 0.000), with 98.11% of the injured muscles being correctly classified (cut-off point 12.50% on Tc differences), and an AUPRC value of 0.981, with association classification criteria at >9.87. Tensiomyography has a high predictive ability to discriminate between injured and non-injured BF non-invasively and functionally.