American Academy of Sports Physical Therapy Platform Presentation Abstracts SPL1–SPL82

Abstract

Previous article Next article Full AccessAPTA Combined Sections MeetingAmerican Academy of Sports Physical Therapy Platform Presentation Abstracts (SPL1–SPL82)AuthorsJournal of Orthopaedic & Sports Physical TherapyPublished Online:December 31, 2020Volume51Issue1PagesCSM24-CSM56https://www.jospt.org/doi/10.2519/jospt.2021.51.1.CSM24SectionsPDFPDF PLUSAbstract ToolsAdd to FavoritesDownload CitationsTrack CitationsPermissions ShareShare onFacebookTwitterLinkedInRedditEmail AboutAbstractThese abstracts are presented here as prepared by the authors. The accuracy and content of each abstract remain the responsibility of the authors. In the identification number above each abstract, SPL designates an American Academy of Sports Physical Therapy platform presentation.J Orthop Sports Phys Ther 2021;51(1):CSM24–CSM56. doi:10.2519/jospt.2021.51.1.CSM24SPL1US Swimmers' Shoulder Pain and Coaches' Experience, Education, and Views on Injury PreventionLisa Baek, Brian David Stirling, Jonathan C. Sum, Adam S. Barrack, Lori Michener, Angela Rosemarie TatePURPOSE/HYPOTHESIS: Shoulder pain is common in competitive swimmers. Age, training volume, and lack of dryland training have been cited as factors, but coaches' training and experience have not been investigated. The purposes were to (1) compare current pain, disability and training volume among 4 competitive age groups (9–10, 11–12, 13–14, 15–17 years) in USA swimming, (2) quantify coaches' experience, education and views on shoulder pain and prevention, and (3) determine the effect of coaching experience on training, shoulder pain, and disability.NUMBER OF SUBJECTS: Four hundred fifty-two swimmers and 20 coaches.MATERIALS AND METHODS: Six-state multicenter study used separate surveys for swimmers and coaches. Swimmers completed a Disability of Arm Shoulder Hand sports module and Penn Pain Score. Coaches reported years of coaching experience, level of American Swimming Coaches Association (ASCA) training, estimates of swimmers' pain prevalence, preferred methods of injury prevention, and training parameters. ANOVAs with post hoc comparisons were performed to determine effects of age, coaching experience defined by quartiles, swimming yardage, dryland training hours, disability, and pain.RESULTS: The older age groups (13–14, 15–17) swam significantly more yardage and hours than younger age groups (P<.001). Swimmers aged 15 to 17 had more pain than 11 to 12 (P = .003) and more pain (P<.001) and disability (P = .002, Bonferroni-corrected α = .007) than 9 to 10. More experienced coaches (greater than 11 years) prescribed greater swimming yardage and dryland training hours (P<.017, Bonferroni-corrected α = .017), but there were no differences in pain or disability. 87.3% coaches reported additional coaching education, 71.8% reported better physical conditioning and 16.9% reported rule modification to constrain yardage would reduce shoulder injury. 81.7% coaches estimated that 10% or less of their swimmers had pain. 50.2% of athletes reported current pain with swimming.CONCLUSIONS: Older swimmers incurred greater pain and disability and swam more yardage than younger swimmers, yet few coaches felt that yardage constraint would reduce shoulder injury. Potential factors mediating the effects of loading, such as dryland training, require further investigation. Sixty-nine percent of swimmers had coaches who had not completed ASCA level 3 module on the physiology of training, which requires coaches' swimmers to meet performance criteria for them to qualify. Given that 87% of coaches reported additional education would reduce injury, a lack of access may explain the low level of coaching education.CLINICAL RELEVANCE: Given the increasing pain and disability among older swimmers, injury prevention programs are warranted and potential factors mediating shoulder pain should be investigated. Consideration should be given to providing education to coaches without regard for their swimmers' performance. A vast disparity exists among coaches' estimates and swimmers' actual pain prevalence.SPL2Does Intraoperative Platelet-Rich Plasma Enhance Psychological Readiness and Clinical Outcomes Following Anterior Cruciate Ligament Reconstruction With Meniscus Repair?Lane Bailey, Adam O'brien, Jacquelyn Lee Kleihege, Ryan Warth, Michael Sehlinger, Brian Randall Duncan, Walter R. LowePURPOSE/HYPOTHESIS: Orthopaedic surgeons have given recent attention to use of ortho-biologics to potentially augment the biologic healing timelines after anterior cruciate ligament reconstruction (ACLR) with meniscal repair. Specifically, platelet-rich plasma (PRP) holds potential benefits to enhance the natural healing response but has shown mixed results with regard to measurable clinical benefit. The purpose of this study was to compare psychological readiness and functional outcomes of patients undergoing ACLR with meniscal repair between those who received concomitant intraoperative PRP versus age, gender and graft matched controls. We hypothesized that patients receiving intraoperative PRP would exhibit higher psychological readiness scores at return to sport and better functional outcomes at 2 years when compared to matched controls.NUMBER OF SUBJECTS: Six hundred seventy-three patients undergoing primary ACLR with meniscal repair from a single surgeon.MATERIALS AND METHODS: One hundred sixty-two subjects who received intraoperative PRP were age, sex, and graft-type matched to a control group (n = 162). Baseline data were gathered for patient and surgical demographics including preinjury level of activity (MARX scale), meniscus injury size/location, and mean adjusted gross income (AGI). Psychological readiness was compared at the time of return to sport using the Anterior Cruciate Ligament-Return To Sport After Injury scale (ACL-RSI). At 2 years groups were compared for reinjury rates and subjective knee function (SANE score). Data were assessed for normality using Shapiro-Wilk test and comparisons were analyzed using univariate generalized linear models with an alpha level of .05.RESULTS: There were no baseline differences with regards to BMI (P = .72), MARX score (P = .82), meniscus injury size (P = .83) and location (P = .59). Additionally, no differences were observed for mean AGI between groups (less than $50k/year, P = .95; $50k-$100k/year, P = .37; greater than $100k/year, P = .77). At return to sport, there were no mean differences in ACL-RSI scores between the PRP and control groups, respectively (90.7 versus 87.3, P = .10). At 2 years, there were no differences in the reinjury rate (3.2% versus 3.0%, P = .73) or SANE self-reported knee function score (91.6 ± 11.2 versus 92.4 ± 10.6, P = .60) for those receiving PRP compared to the matched controls.CONCLUSIONS: The added use of intraoperative PRP has no measurable benefit to psychological readiness at return to sport following ACLR with meniscal repair. Additionally, the use of intraoperative PRP does not appear to influence reinjury rates, and self-reported knee function at 2 years.CLINICAL RELEVANCE: This study suggests that while orthobiologics have promising basic science and theoretic uses, further investigation on the clinical benefits are warranted before considering this modality as a potential standard of care.SPL3Baseball Pitcher–Specific Elbow Torque-Velocity Relationship: Influence of Physical FactorsAdam S. Barrack, Holly Rittenberry, Alia Yasen, Nicholas Lobb, Ryan Zerega, Roksana Karim, Lori Michener, Andrew KardunaPURPOSE/HYPOTHESIS: Elbow valgus torque is a surrogate for assessing ulnar collateral ligament injury (UCL) risk, as the UCL provides 54% resistance to elbow valgus torque. Higher ball velocity is associated with greater elbow valgus torque in baseball pitchers; however, this relationship is not consistent across pitchers. Physical factors may mediate the torque-velocity (T-V) relationship. Here we characterize the pitcher-specific T-V relationship, and the influence of physical factors on the relationship.NUMBER OF SUBJECTS: Twenty-seven NCAA Division I college pitchers (mean ± SD age, 20.4 ± 1.4 years; height, 1.9 ± 0.05 m; weight, 90.9 ± 7.8 kg).MATERIALS AND METHODS: Shoulder and hip isometric strength, shoulder and hip range of motion (ROM), and lumbopelvic stability were measured using a handheld dynamometer, inclinometer, and a unilateral hip bridge test, respectively. Next, pitchers threw a 10-pitch bullpen while peak elbow valgus torque was recorded using a Motus Baseball IMU taped directly to the skin at the elbow, and ball velocity was recorded with a radar gun (Stalker Radar). Pitchers threw 5 fastballs at 75% effort and 5 at 100% effort to ensure variation in pitch velocity. These data were used in a linear regression to characterize the T-V relationship, and to determine the mediating relationship of the physical factors on the T-V relationship.RESULTS: Pitch velocity significantly predicted variability in elbow torque in 13 of 27 pitchers (SigT-V) average R2 = 0.72; range, 0.47–0.90; P<.05). The remaining 14 pitchers did not have a significant torque-velocity relationship (Non-SigT-V) (average R2 = 0.15; range, 0.001–0.33; P>.05). The SigT-V group exhibited significantly (P<.15) greater shoulder internal rotation strength rate of force development (P = .08) and less total hip range of motion (external + internal rotation; TROM) (P = .10) than the Non-SigT-V group. Both physical measures mediated the T-V relationship; shoulder internal rotation strength rate of force development (increase in R2 = 0.02) and hip TROM (increase in R2 = 0.01).CONCLUSIONS: The relationship between torque and velocity is pitcher-specific, with 13 pitchers demonstrating a significant relationship. Two physical factors mediated the T-V relationship. Decreased rate of force development during shoulder internal rotation in the Non-SigT-V group indicates that increased acceleration at the shoulder may mediate elbow valgus torque. Greater hip TROM as seen in the Non-SigT-V group indicates that hip ROM may mitigate valgus torque about the elbow during pitching.CLINICAL RELEVANCE: Ball velocity alone cannot consistently predict arm stress, as this relationship is pitcher-specific. Using modifiable physical measures in addition to torque and ball velocity may provide a more accurate assessment of pitcher UCL injury risk. This player-specific injury risk profile can also be used to guide rehabilitation and return to sport.SPL4Return-To-Sport Strength Symmetry Impacts 5-Year Cartilage Integrity After Anterior Cruciate Ligament ReconstructionCaroline Lewis Brunst, Matthew Peter Ithurburn, Andrew Zbojniewicz, Mark Vincent Paterno, Laura C. SchmittPURPOSE/HYPOTHESIS: After ACL reconstruction (ACLR), quadriceps femoris (QF) strength is a critical component of return-to-sport (RTS) decisions, as deficits that persist after rehabilitation contribute to worse function and altered movement. QF weakness is also implicated in early development of knee osteoarthritis; however, further understanding of the impact of QF strength at the time of RTS on longitudinal outcomes is needed. This study investigated the impact of QF strength at the time of RTS after ACLR on knee cartilage integrity 5 years later. The primary hypothesis tested was that participants with QF asymmetry at RTS would demonstrate magnetic resonance imaging (MRI) markers of knee cartilage degeneration measured at 5 years post RTS.NUMBER OF SUBJECTS: Twenty-seven young athletes after primary, unilateral ACLR (mean ± SD age, 16.8 ± 2.3 years at the time of RTS; 22 female).MATERIALS AND METHODS: Participants were enrolled at medical clearance for RTS (not controlled by the study) and followed for 5 years. At RTS, isometric QF strength was measured bilaterally (isokinetic dynamometry) and limb symmetry index (LSI) was calculated. QF LSI was used to divide the cohort into 2 groups: High QF (RTS-HQ; LSI of 90% or greater) and Low QF (RTS-LQ; LSI less than 85%) based on published work. At 5 years post RTS, quantitative MRI sequences were performed on the involved knees. T2 mapping (relaxation times in msec; higher values indicative of cartilage degeneration) was performed for the medial/lateral femoral condyle and medial/lateral tibia. Group differences in T2 relaxation times were compared between the RTS-HQ and RTS-LQ groups with independent-samples t tests (α≤.05).RESULTS: The RTS-HQ (n = 13) and the RTS-LQ (n = 14) groups did not differ in age (at the time of RTS; 16.9 versus 16.7 years, respectively; P = .79), sex distribution (33% versus 48% female, respectively; P = .17), or time from ACLR to RTS clearance (8.4 versus 7.1 months, respectively; P = .15). The RTS-LQ group demonstrated higher T2 relaxation times in the superficial layers of the lateral femoral condyle in the anterior region (40.0 versus 34.3 milliseconds, respectively; P = .007) and middle region (35.9 versus 29.6 milliseconds, respectively; P = .008) compared to the RTS-HQ group, with a trend towards elevated full-thickness relaxation times in the anterior region (37.5 versus 33.3 milliseconds, respectively; P = .054).CONCLUSIONS: Markers of knee cartilage degeneration were observed in young athletes 5 years post RTS after ACLR. Specifically, young athletes with QF strength asymmetry at the time of RTS after ACLR demonstrated elevated knee cartilage T2 relaxation times, indicative of cartilage degeneration, in the lateral femoral compartment 5 years later.CLINICAL RELEVANCE: Just over 50% of this cohort was cleared for RTS with involved limb QF strength deficits that may contribute to early markers of knee cartilage degeneration within the subsequent 5 years. While investigation with larger samples and consideration of confounding factors is needed, this work highlights the need to restore QF strength symmetry during rehabilitation for young athletes after ACLR to mitigate risk of early cartilage degeneration.SPL5Do Lower Extremity Asymmetries Impact Utilization Rates of the Athletic Training Room in Division I Athletes?Hannah Christa Buckingham, Shefali M. Christopher, Srikant Vallabhajosula, Mason Lee MoorePURPOSE/HYPOTHESIS: Previous studies have identified lower extremity asymmetries as a risk factor for injury, however, little information exists on whether these asymmetries impact how much an athlete uses the athletic training room (ATR). The aim of our study was to determine if athletes with lower extremity asymmetries utilize the ATR more than those who are more symmetrical at baseline functional testing.NUMBER OF SUBJECTS: Seventy-six.MATERIALS AND METHODS: Preseason functional testing of the Single Leg Hop Test (SLH) and Lower Quarter Y Balance Test (LQYBT), were administered prior to the 2019–2020 competition year for Division I athletes at 1 university. The football, men's and women's soccer, men's and women's cross-country, women's lacrosse, and women's tennis teams were tested. Asymmetry values for the SLH (defined as absolute value of 100 [percent ratio of left and right leg hop distance]) and LQYBT (defined by absolute difference between left and right reach distance in anterior, posterolateral [PL] and posteromedial [PM] directions), injury status and ATR utilization rates were used to run correlation analysis. Baseline SLH and LQYBT performance was also compared between those who were injured and uninjured using Mann-Whitney U test.RESULTS: Seventy-six athletes were tested at baseline. Of those, 32 had injuries throughout the competition year (shortened due to COVID-19). There was a significant weak negative relationship between the SLH (P = .041, r = −0.27), LQYBT PL direction (P = .044, r = −0.25) and total ATR utilization rates (URT). For athletes that were injured during the season, there was a significant moderate negative relationship between URT and LQYBT PL (P = .007, r = −0.494). There was a significant weak negative relationship between the LQYBT Anterior Reach and prevention utilization rate (P = .048, r = −0.315) for uninjured athletes. No significant differences were found between ATR utilization rates between injured and uninjured groups. Significant differences were noted between injured (3.68 cm) and uninjured groups (5.41 cm) for the LQYBT PM reach distance (P = .030).CONCLUSIONS: Athletes with greater asymmetry at baseline on the SLH and LQYBT PL utilized the ATR less throughout the competition year than those with less asymmetry. This was also true for athletes that got injured during the competition year. Although statistically significant, these correlations were weak to moderate and more research is needed to quantify the relationship of ATR utilization and baseline functional testing.CLINICAL RELEVANCE: Traditionally, athletes have used the ATR when injured, however, recent literature indicates athletes utilize the ATR the most for preventive measures and non- time loss injuries. Our study may indicate athletes with greater asymmetry at baseline may underutilize the ATR, necessitating encouragement/education to use the ATR more for preventative purposes. Future studies should explore why athletes with less asymmetry or injury utilize the training room so that allocation of time, resources, and education of staff can meet the current needs of the athletes.SPL6Development and Internal Validation of a Humeral Torsion Prediction Model for Professional Baseball PitchersGarrett Scott Bullock, Ellen Shanley, Gary S. Collins, Nigel K. Arden, Amanda Jo Arnold, Thomas Noonan, Micheal Kissenberth, Lane Bailey, Charles Alden ThigpenPURPOSE/HYPOTHESIS: Baseball throwing generates high torque through the arm which influences osseous humeral adaptation, termed humeral retrotorsion (HT). HT has been linked to pitching arm injury risk and influences shoulder range of motion. However, clinical HT assessment requires expensive equipment which limits use. Developing a HT predictive model using standard clinical tests and measures may aid clinical application. Therefore, the purpose of this study was to develop and internally validate a HT prediction model using standard clinical tests and measures in a sample of professional baseball pitchers.NUMBER OF SUBJECTS: Four hundred seven professional baseball pitchers.MATERIALS AND METHODS: An 11-year (2009–2019) prospective professional baseball cohort was used for this study. Participants were included if they were able to participate in all practices and competitions. Preseason shoulder range of motion (External [ER], Internal [IR], Horizontal Adduction [HA]) and HT were collected each season. Player age, arm dominance, arm injury history and continent of origin were also collected. Examiners were blinded to arm dominance. An a priori power analysis determined that 244 players were needed for accurate prediction models. Missing data were low (less than 3%), thus a complete case analysis was performed. Model development followed the transparent reporting of a multivariable prediction model for individual prognosis or diagnosis (TRIPOD) recommendations. Regression models with restricted cubic splines were performed. Following primary model development, bootstrapping with 2000 iterations were performed to reduce overfitting and assess optimism shrinkage. Prediction model performance was assessed through root mean square error (RMSE), R2 and calibration slope with 95% confidence intervals (CI). Sensitivity analyses included dominant and nondominant HT.RESULTS: Demographics included age (23.2 ± 2.4 years), BMI (25.1 ± 2.3 kg/m2), Left Handed (17%), Arm History Prevalence (21%), HT (11 ± 14), IR (41° ± 12°), ER (126° ± 13°), HA (19° ± 16°). Final model RMSE was 11.86, R2 was 0.41, and calibration was 1.00 (95% CI: 0.94, 1.06). For each degree increase in IR explained a 0.37° (95% CI: 0.31°, 0.43°; P<.001) of HT (95% CI: 0.31, 0.43; P<.001). Every degree increase in ER explained −0.30° (95% CI: −0.35°, −0.24°; P<.001) of HT. Every degree increase in HA explained 0.31° (95% CI: 0.26°, 0.36°; P<.001) of HT. Right handed pitchers demonstrated −1.86° (95% CI: −3.58°, −0.14°; P = .034) HT compared to left handed pitchers. Sensitivity analyses demonstrated similar model performance.CONCLUSIONS: This is the first study to show that clinical ROM measurements can be used to accurately predict HT in professional baseball pitchers. This provides a reasonable method to include HT in clinical decisions regarding shoulder ROM and arm injury risk identification. Futures studies should externally validate this approach.CLINICAL RELEVANCE: This predictive model can be used by clinicians to infer HT using standard clinical test and measures. These data should be considered when evaluating professional pitchers for arm injury risk.SPL7Machine Learning is Not Superior To a Traditional Statistical Prediction Model of Humeral TorsionGarrett Scott Bullock, Ellen Shanley, Gary S. Collins, Nigel K. Arden, Thomas Noonan, Micheal Kissenberth, Amanda Jo Arnold, Lane Bailey, Charles Alden ThigpenPURPOSE/HYPOTHESIS: Baseball throwing generates high arm forces which influence humeral osseous adaptation, termed humeral retrotorsion (HT). HT has been associated with arm injuries and shoulder range of motion. Currently HT measurement requires expensive tools, decreasing clinical utility. Machine learning (ML) is a novel strategy used to predict outcomes with high complexity. Comparing statistical and ML techniques allows for the best prediction model to be clinically implemented. Therefore, the purpose of this study was to develop ML models to predict youth baseball pitcher HT and compare them to a regression prediction model.NUMBER OF SUBJECTS: Four hundred twenty-three youth baseball pitchers.MATERIALS AND METHODS: A 2-year prospective cohort was conducted in young baseball players. Participants were included if (1) pitcher was their primary position (2) able to participate in all practices and competitions, (3) ages 8 to 18. Preseason shoulder range of motion (External [ER], Internal [IR], Horizontal Adduction [HA]) and HT were collected. Player age, arm dominance, arm injury history, and parental height and weight were also collected. Biologic maturity age was calculated using the Khamis-Roche protocol. z scores were calculated as the difference between biological maturity age and chronological age. A priori power analysis determined that 406 players were needed for accurate prediction models. Missing data were 3% or less, thus a complete case analysis was performed. A regression model with restricted cubic splines was performed. ML models included: Random Forest, Gradient Boosting Machine, Support Vector Machine Regression, and Artificial Neural Networks. ML models were calculated for best hyperparameter tuning. Ten-fold cross-validation was performed for regression and ML models. Model performance was assessed through root mean square error (RMSE) and calibration slope with 95% confidence intervals.RESULTS: Demographics included age (11.7 ± 2.0 years), BMI (27.5 ± 5.5 kg/m2), Left Handed (13%), Biologic Maturity z score (0.28 ± 0.81), Arm History Prevalence (21%), HT (11 ± 14), IR (41° ± 12°), ER (126° ± 13°), HA (19° ± 16°). Regression RMSE was 13.64 and calibration was 1.00 (0.89–1.11). Random Forest RMSE was 7.57 and calibration was 1.49 (1.44–1.53). Gradient boosting machine RMSE was 10.23 and calibration was 1.28 (1.21–1.35). Support vector machine RMSE was 9.68 and calibration was 1.35 (1.24–1.38). Artificial neural network RMSE was 11.30 and calibration was 1.00 (0.91–1.09).CONCLUSIONS: This is the first study to show that ML does not improve youth baseball HT prediction compared to a traditional regression model. This suggests traditional regression models should be used for HT prediction. While ML models demonstrated improved RMSE it displayed poorer calibration compared to the statistical model.CLINICAL RELEVANCE: Based on these results we recommend a simple equation from a statistical model which can be quickly and efficiency integrated within a clinical setting. This prediction model uses standard clinical tests measures which aid in improving youth baseball examination.SPL8Application of Machine Learning to Improve Vertical Ground Reaction Force Prediction Accuracy Using Inertial SensorsCourtney Ross Chaaban, Nathaniel T. Berry, Cortney Armitano-Lago, M. Spencer Cain, Adam Kiefer, Michael J. Mazzoleni, Erik Wikstrom, Darin PaduaPURPOSE/HYPOTHESIS: Vertical ground reaction force (vGRF) is a measure of limb loading that is commonly reported when assessing injury risk and readiness for return to sport. However, most clinicians lack access to the force plates required to obtain this measure. Inertial measurement units (IMUs) are a practical alternative to measure lower extremity loading in both clinical and real-world environments. While previous research has shown moderate correlations between tibial acceleration and vGRF for individual subjects using person-specific models, we sought to develop a prediction model utilizing multiple IMU features that avoided person-specific parameters and was robust enough to apply to multiple subjects. Therefore, the purpose of our study was to assess the capability of thigh- and shank-mounted IMUs to predict peak vGRF in each limb during a double limb landing task utilizing a robust data-driven methodology.NUMBER OF SUBJECTS: Sixteen healthy subjects (14 female; mean ± SD age, 20.6 ± 1.2 years; mass, 67.9 ± 10.9 kg; height, 170 ± 7 cm) participated in this study.MATERIALS AND METHODS: IMUs with triaxial accelerometers and gyearoscopes were secured to the proximal tibia and distal femur of each limb. Subjects completed 8 trials of a double limb drop vertical jump task with concurrent collection of 3-D marker and force data. Both single and multiple-feature models were used to predict peak vGRF normalized to body weight from inertial data. First, a simple linear regression model was used to predict peak vGRF from the max-min difference in axial tibial acceleration. Next, 2 separate models were used to predict peak vGRF: one using features derived from the accelerometer data only (ACCEL) and another one using features derived from both the accelerometer and gyearoscope data (ACCEL_GYRO). Thirty total features were extracted from the 4 accelerometer signals and an additional 40 total features were extracted from the 4 gyearoscope signals and input into the initial model. These features were selected based on their biomechanical relevance and ease of calculation (eg, positive and negative peak values, time to peak values). Left and right limbs were analyzed together (n = 256) using 10-fold cross-validation. Final model selection was performed using stepwise linear regression and the Akaike information criterion, with 12 features being selected for the ACCEL model and 22 features being selected for the ACCEL_GYRO model.RESULTS: The simple linear regression was statistically significant (R2 = 0.22, P<.001) with an RMSE of 0.54 times body weight. The ACCEL model was statistically significant (R2 = 0.74, P<.001) with an RMSE of 0.32 times body weight. The ACCEL_GYRO model was statistically significant (R2 = 0.82, P<.001) with an RMSE of 0.26 times body weight.CONCLUSIONS: Our findings suggest that IMUs can predict vGRF during drop landing tasks. We found improved prediction accuracy with the utilization of multiple sensors and multiple features. We utilized variable reduction and cross-validation methods to avoid overfitting our models.CLINICAL RELEVANCE: Given that IMUs are inexpensive and portable, we suggest clinicians can use IMUs to quantify lower extremity loading during jump landing tasks.SPL9Neural Activity for Uninvolved Knee Motor Control in Individuals Following Anterior Cruciate Ligament ReconstructionMeredith Chaput, Cody R. Criss, Janet E. Simon, James A. Onate, Dustin R. GroomsPURPOSE/HYPOTHESIS: Anterior cruciate ligament (ACL) injury induces neurophysiologic changes for motor control of the knee joint which continues to impair patient function despite reconstructive surgery (ACLR) and rehabilitation. Previous functional magnetic resonance imaging (fMRI) studies have focused on neural activation patterns for the involved limb. However, sensorimotor deficits occur bilaterally, but the neurophysiologic mechanisms for uninvolved knee deficits are not well understood. Thus, our purpose was to investigate neuroplasticity after ACLR that occurs with uninvolved limb movement relative to controls. We hypothesized that the ACLR group would have increased neural activity in brain regions responsible for cognitive control of movement.NUMBER OF SUBJECTS: Fifteen patients with left (involved) knee ACLR (8 female; mean ± SD age, 21.53 ± 2.7 years; height, 173.22 ± 10.0 cm; weight, 72.15 ± 16.1 kg; Tegner activity level, 7.40 ± 1.1; 43.33 ± 33.1 months post surgery; 2 patellar tendon and 13 hamstring grafts) and 15 age, sex, weight, height, sport, activity level, and education level matched controls (8 female; age, 23.33 ± 2.7 years; height, 174.92 ± 9.7 cm; weight, 72.14 ± 15.4 kg; Tegner activity level, 7.33 ± 1.0) participated.MATERIALS AND METHODS: fMRI was completed on a 3-T Siemens Magnetom scanner with a 16-channel head coil. Brain imaging to capture neural activity (quantified via Blood Oxygen Level Dependent [BOLD] signal) during four 30-second cycles of a right (uninvolved) knee flexion/extension task paced with a metronome was completed. Brain activation patterns during knee flexion/extension were contrasted with an interspersed rest condition. Each subject's data were averaged across trials. Results are reported as z scores, with higher scores indicating higher likelihood of activation and P value indicating probability of finding activation when there is none. The ACLR and control groups were compared using a general linear model and z-threshold greater than 2.3.RESULTS: The ACLR group had increased BOLD signal in one statistical cluster in regions corresponding