Kinematics-Based Recovery Metrics and Inertial Measurement Units to Monitor Recovery Post-Knee Arthroscopic Surgery: A Case Study

Physiotherapy after lower-limb injury or surgery is essential for recovery of range of motion, functional movement, strength, and return to sport. Clinicians assess patients, prescribe rehabilitation exercises, and monitor progress through recovery phases. Given the bulk of recovery occurs between in-person visits, coupled with regional differences in access to physiotherapy care, remote monitoring of recovery is warranted to improve patient care and recovery. \nThis work follows the recovery of a patient after arthroscopic partial meniscectomy (APM) surgery, a procedure to remove part of the meniscus in the knee joint. The meniscus is a tissue in the knee joint that improves the articulating surface between the femur and tibia, shock absorption, and transmits force. A conservative estimate puts the rate of meniscal tears at 60 per 100,000, making the APM procedure one of the most common orthopaedic procedures performed. Rehabilitation after APM procedure is generally separated into three phases where the continuation to the next phase relies on meeting the goals \nof the previous phase as determined by clinician assessments. Assessments are often done through visual observation, manual testing, and goniometric measurements. In a remote setting, these assessments and measurements are challenging to conduct. Wearable inertial measurement units (IMUs) can reconstruct 3D human motion in an unconstrained space, making them a potentially useful tool for remote visualization of therapy exercises and for generating recovery metrics that clinicians can use to inform decision making. \nThe first part of this work extracts current and exploratory recovery metrics to examine recovery over time, alignment with clinical decisions, and explores novel metrics quantify recovery remotely. Exploratory recovery metrics were extracted based on literature review, clinical input, and incidental findings. Fifty-one (51) recovery metrics were extracted for 5 of the most common rehabilitation exercises: supine heel slide, leg raise, straight line walking, goblet squats, and single leg Romanian deadlifts. Metrics showed strong evidence of recovery if all of the following conditions were observed: improving trends over the recovery period, trends between affected and unaffected limbs, and significant differences. Metrics showed moderate evidence of recovery if two of three conditions were met and weak evidence of recovery if only one or no conditions were met. Of all the metrics examined, 39.2% (20/51) of metrics provided strong evidence of recovery, determined by trends over \nrecovery, between affected and unaffected limbs, and statistical significance. An additional 45.1% (22) of the metrics showed moderate evidence of tracking recovery over time for this case study. Of the 23 exploratory recovery metrics examined, 13 showed strong evidence of recovery and potential for use in tracking rehabilitation. \nThe second component of this thesis examined the IMU metric error relative to motion capture-based metrics and exercise specific tuning of the IMU algorithm noise parameters. Error between IMU and motion capture metrics being smaller than the effect size, as well as IMU metrics demonstrating similar recovery trends to motion capture metrics, were factors considered when determining the remote monitoring potential using IMU metrics. IMU feasibility was considered strong if both these conditions were met, moderate if only one condition was met, and weak if neither condition was met. Fourteen (14) metrics showed strong feasibility for remote monitoring using the algorithm and another 24 metrics showed moderate feasibility. Tuning the IMU algorithm measurement noise parameters for the heel slide and leg raise showed that increasing gyroscope noise improved heel slide metric error 9.48%, while decreasing gyroscope noise improved metric error for the leg raise exercise by \n23.5%. \nFinally, a clinician survey was conducted to gather clinician feedback on recovery metrics and stakeholder opinion on future use of the data. As the target primary users of the data presented in this work, 19 physiotherapists participated in the survey. For all metrics they currently use, 95.5% of respondents said they would use the data provided to assist in monitoring recovery. Eight-one percent (81.1%) of respondents said they would potentially use data from exploratory recovery metrics to assist in their clinical decision making, if the data was available. Strength of clinician feedback from the survey was based on the percentage of responses that said they would use the data to inform therapeutic decision making. \nThis work presents examination of new and existing recovery metrics and a wearable IMU system to monitor recovery remotely using a case study of a patient recovery from a lower limb surgery. Existing metrics provide good indication of recovery, while a subset of exploratory metrics show potential to add valuable recovery information given further validation. Preliminary results indicate that setting exercise specific tuning parameters might have potential for better algorithm performance. Initial clinician feedback on motion capture metrics and future use was primarily positive. Overall, 10 metrics are rated as strong in all two or three categories. Six (6) other metrics were tracked well using the IMU algorithm, however did not show recovery in this case study. Ten (10) metrics showed trends over the recovery period, but only demonstrated moderate success tracking trends using IMUs. Combined, the information presented in this work shows promise in improving patient care and recovery, potentially increasing access to quality care, and transitioning sensor-based human movement reconstruction tools to a clinical setting.

Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.