Objectively Differentiating Anterior Cruciate Ligament Injury Status in Adolescents Through Clinical and Laboratory-Based Measures

Abstract

Background: Anterior cruciate ligament (ACL) injuries in adolescents are increasingly common and carry long-term consequences, including reduced return-to-activity (RTA) rates, elevated reinjury risk, and early-onset osteoarthritis. Current RTA assessments rely heavily on subjective questionnaires, strength measures, and spatiotemporal outcomes from a variety of functional tasks. While these tools are easy to administer, they often fail to capture the quality of the movement, leaving subtle but important deficits undetected. As a result, many athletes satisfy current clinical criteria despite exhibiting movement patterns that differ from those of healthy peers, which may contribute to persistent functional deficits and potentially elevated reinjury risk. There is a pressing need for approaches that move beyond simple distance- or time-based measures to capture the complexity of adolescent movement patterns and provide clinically relevant, portable solutions. Purpose: This thesis aimed to evaluate whether adolescent functional performance following ACL injury can be more effectively characterized through kinematic analyses that extend beyond conventional spatiotemporal metrics. Four interrelated studies were conducted to determine whether hop tests are reproducible in adolescents, whether spatiotemporal outcomes alone provide sufficient classification accuracy, whether waveform-based analyses can identify movement adaptations not captured by conventional performance measures, and whether simulated inertial measurement units (sIMU) offer a clinically translatable alternative to motion capture (MoCap). Methods: Across four studies, 170 adolescents completed a battery of single-limb hop tasks. The first study assessed the reproducibility and responsiveness of hop tasks in adolescents across environments, using test-retest analyses in a laboratory and gymnasium setting. This study also examined whether spatiotemporal outcomes from hop tasks could differentiate performance between sexes. The second study evaluated whether spatiotemporal variables from these hop tasks were sufficient for classifying ACL injury status. Decision tree models were employed to determine the most discriminative tasks and parameters, and classification outcomes were compared between males and females. The third study shifted focus from spatiotemporal outcomes to continuous kinematic data, applying principal component analysis (PCA) to waveform features extracted from hop tasks. This study tested whether waveform-derived features provided superior classification of ACL injury status compared to spatiotemporal metrics and explored potential differences between injured and contralateral limbs, with secondary exploratory analyses of sex effects. The fourth study extended these findings into a translational context by generating sIMU signals from motion capture trajectories. Multiple sensor configurations were simulated to evaluate whether sIMUs could replicate MoCap-based classification accuracy and to identify which sensor placements provided the most discriminative information for detecting ACL injury status. Results/Discussion: Hop tasks demonstrated reproducibility in adolescents, confirming their suitability beyond laboratory conditions, with sex differences evident in hop distance and symmetry. Spatiotemporal variables provided modest classification accuracy, particularly in females, highlighting their limitations. Waveform-derived PCA features improved classification and revealed adaptations between injured and contralateral limbs, with exploratory analyses suggesting sex-related misclassification patterns. Finally, sIMUs demonstrated moderate classification performance, with results approaching those of MoCap in tasks characterized by larger dynamic movement signatures, with the thigh and shank placements providing the strongest discriminative power. Together, these findings demonstrate that while spatiotemporal outcomes remain useful, waveform and wearable approaches provide deeper insights and greater clinical translatability. Conclusion: This thesis establishes that adolescent ACL injury assessment benefits from moving beyond traditional spatiotemporal outcomes toward kinematic waveform analysis and wearable sensor technologies. By confirming reproducibility, identifying the limitations of existing measures, advancing analytic techniques, and demonstrating the feasibility of sIMUs, this work builds a continuum from controlled laboratory testing to practical field-based applications. These findings contribute to developing more robust, ecologically valid assessments that may ultimately improve rehabilitation monitoring and reduce reinjury risk in adolescent athletes.