Krithika Swaminathan

All Publications

• Quantifying Anterior Cruciate Ligament Injury Resilience: A Screening and Composite Score Framework. Orthopaedic journal of sports medicine Seagers, K., Swaminathan, K., Kolesar, J., Hamner, S., Muccini, J., Hicks, J., Myer, G., Hastie, T., Delp, S. 2026; 14 (5): 23259671261433009

  Abstract

  Over 1.4 million anterior cruciate ligament (ACL) injuries occur worldwide each year, but studies suggest that they may be preventable through targeted screening and training. Biomechanical factors from squatting, jumping, cutting, and running have been associated with ACL injury risk; however, current movement assessments are lengthy, limiting adoption and compliance. There is no consensus on which activities best assess modifiable biomechanical risk factors.(1) To identify an optimal set of activities that explain an athlete's full-body, 3-dimensional biomechanical risk factors for ACL injury and (2) to build a framework to determine an individual's ACL injury resilience (AIR) score.Descriptive laboratory study.The authors recruited adolescent female athletes who played soccer, basketball, or volleyball as their primary sport and had no history of injury. Data collection included 3-dimensional motion capture and ground-reaction forces during 5 common ACL injury risk screening activities and the computation of 35 biomechanical risk factors. Column subset selection identified the set of activities that best reconstructed the biomechanical factors from left-out activities. This activity set, with biomechanical thresholds from prospective studies of ACL injuries, was used to build the AIR score framework. The AIR score and the subscores for the trunk, hip, knee, and foot ranked participant biomechanical resilience. Statistical significance was evaluated at α = .050.Twenty-seven females were included in this study (mean ± SD; age, 15.7 ± 1.2 years; body mass index, 22.19 ± 3.58 kg/m2). Two screening activities-run cuts and single-leg drop jumps-achieved the highest reconstruction performance, predicting the biomechanical factors from the left-out activities with 88.4% accuracy. The AIR score from this 2-activity set strongly agreed with an AIR score generated using all the information from 5 activities (r = 0.90; τ = 0.72; P < .001).The AIR screen and score provided a robust methodological and statistical framework that could be updated as future studies identify and characterize new biomechanical risk factors in larger cohorts.The authors' proposed screening provided an individualized evaluation of ACL injury resilience from just 2 activities, using an objective assessment of 3-dimensional movement. The AIR score framework identified athletes with vulnerable whole-body movement patterns, along with subscores that could be targeted with exercise interventions.

  View details for DOI 10.1177/23259671261433009

  View details for PubMedID 42221210

  View details for PubMedCentralID PMC13219899

• Learning-based 3D human kinematics estimation using behavioral constraints from activity classification NATURE COMMUNICATIONS Kim, D., Jin, Y., Cho, H., Jones, T., Zhou, Y., Fadaie, A., Popov, D., Swaminathan, K., Walsh, C. J. 2025; 16 (1): 3454

  Abstract

  Inertial measurement units offer a cost-effective, portable alternative to lab-based motion capture systems. However, measuring joint angles and movement trajectories with inertial measurement units is challenging due to signal drift errors caused by biases and noise, which are amplified by numerical integration. Existing approaches use anatomical constraints to reduce drift but require body parameter measurements. Learning-based approaches show promise but often lack accuracy for broad applications (e.g., strength training). Here, we introduce the Activity-in-the-loop Kinematics Estimator, an end-to-end machine learning model incorporating human behavioral constraints for enhanced kinematics estimation using two inertial measurement units. It integrates activity classification with kinematics estimation, leveraging limited movement patterns during specific activities. In dynamic scenarios, our approach achieved trajectory and shoulder joint angle errors under 0.021 m and 6 . 5 ∘ , respectively, 52% and 17% lower than errors without including activity classification. These results highlight accurate motion tracking with minimal inertial measurement units and domain-specific context.

  View details for DOI 10.1038/s41467-025-58624-6

  View details for Web of Science ID 001465538200001

  View details for PubMedID 40216761

  View details for PubMedCentralID PMC11992036

• Multi-Modal Sensing for Propulsion Estimation in People Post-Stroke Across Speeds IEEE TRANSACTIONS ON NEURAL SYSTEMS AND REHABILITATION ENGINEERING Swaminathan, K., Choe, D. K., Kim, D., Barde, F., Baker, T. C., Wendel, N. C., Chin, A., Bergamo, G., Siviy, C. J., Lee, C., Awad, L. N., Ellis, T. D., Walsh, C. J. 2025; 33: 2273-2285

  Abstract

  Gait rehabilitation is critical for regaining locomotor independence after neuromotor injuries like stroke. Rehabilitation literature indicates the need for such therapy to continue beyond the clinic in order to maintain motor function and support recovery. However, implementing community-based rehabilitation requires the ability to monitor gait in the real-world with clinically relevant accuracies. Despite advances in machine learning, achieving this performance with single sensing modalities has been challenging using wearable sensors like inertial measurement units (IMUs) and pressure insoles. Here, we investigate the benefits of multi-modal sensing by integrating IMU and insole data to develop individualized machine learning models in people post-stroke that estimate propulsion, a key biomechanical variable. We show that in the lab, IMU + Insole models improve performance relative to IMU only and Insole only models, with an average root-mean-squared-error (RMSE) of 0.80 %bodyweight (%BW) across the stance phase. We obtain RMSEs of 0.71%BW for peak paretic propulsion and 0.19%BW s for paretic propulsion impulse, which are within corresponding clinical thresholds. We then explore the application of this algorithm to track propulsion changes in the real-world for two participants during variable-speed walking and two participants during active gait interventions, either functional electrical stimulation or exosuit-applied resistance. For these participants, we observe similar changes in measured propulsion in the lab and estimated propulsion out of the lab across speeds and interventions. Overall, this work aims to address the challenges in applying machine learning methods for individuals post-stroke and presents an investigation into the feasibility of developing estimation methods for real-world propulsion estimation during gait rehabilitation.

  View details for DOI 10.1109/TNSRE.2025.3577961

  View details for Web of Science ID 001508099000004

  View details for PubMedID 40489273

• Estimation of joint torque in dynamic activities using wearable A-mode ultrasound NATURE COMMUNICATIONS Jin, Y., Alvarez, J. T., Suitor, E. L., Swaminathan, K., Chin, A., Civici, U. S., Nuckols, R. W., Howe, R. D., Walsh, C. J. 2024; 15 (1): 5756

  Abstract

  The human body constantly experiences mechanical loading. However, quantifying internal loads within the musculoskeletal system remains challenging, especially during unconstrained dynamic activities. Conventional measures are constrained to laboratory settings, and existing wearable approaches lack muscle specificity or validation during dynamic movement. Here, we present a strategy for estimating corresponding joint torque from muscles with different architectures during various dynamic activities using wearable A-mode ultrasound. We first introduce a method to track changes in muscle thickness using single-element ultrasonic transducers. We then estimate elbow and knee torque with errors less than 7.6% and coefficients of determination (R2) greater than 0.92 during controlled isokinetic contractions. Finally, we demonstrate wearable joint torque estimation during dynamic real-world tasks, including weightlifting, cycling, and both treadmill and outdoor locomotion. The capability to assess joint torque during unconstrained real-world activities can provide new insights into muscle function and movement biomechanics, with potential applications in injury prevention and rehabilitation.

  View details for DOI 10.1038/s41467-024-50038-0

  View details for Web of Science ID 001270732400017

  View details for PubMedID 38982087

  View details for PubMedCentralID PMC11233567