JYJennifer Yu

Jennifer Yu

*This is an algorithmically generated author record based on publicly available information.

Current Organizations / Institutes

The organizations/institutes from the scholar's latest article


Jennifer Yu has a total of 71 co-authors. Here are the top co-authors by number of shared publications:


Latest Publications

  • The statistical robustness of literature on extracorporeal shockwave therapy for the treatment of plantar fasciitis

    by Prabhjot Singh, Mark Kurapatti, Niklas Koehne, Ishandeep Singh, Jennifer Yu, Gray W. Ricca, Nicholas Richardson, Robert L. Parisien

  • Automated scoliosis X-ray cobb angle classification: a deep learning approach with RadImageNet

    by Jennifer Yu, Yash Lahoti, Hamza Ahmed, Mark Kurapatti, Jamie Frost, Junho Song, John Corvi, Nikan Namiri, Tariq Issa, Samuel Cho, et al.

    Study Design Retrospective Cohort Study Objective Biplanar full-body X-ray imaging is a low-dose, high-resolution technology crucial for assessing scoliosis and guiding treatment. Accurate coronal curvature measurement is essential for quantifying severity and planning surgery. Traditional methods are manual, time-consuming, and prone to interobserver variability. Deep learning offers potential for standardization and rapid analysis. RadImageNet, a pretrained model trained on annotated medical images, provides a strong foundation for developing AI-driven radiology tools. This study specifically evaluates whether RadImageNet transfer learning can support adult scoliosis severity classification on standardized AP biplanar full-body radiographs, a task distinct from prior work focused on direct Cobb-angle measurement or adolescent screening. Methods A total of 816 standardized anteroposterior (AP) full-body radiographs from adult patients with degenerative or idiopathic scoliosis and scoliosis-negative cases were used, with an 80/20 split for training/validation and testing, respectively. No external validation cohort was available. Scoliosis severity was classified into three categories based on the highest Cobb angle from the proximal thoracic, main thoracic, and thoracolumbar curves: no scoliosis (0°–10°), mild scoliosis (10°–30°), and severe scoliosis (> 30°). Cobb-angle labels were derived from resident-generated vertebral-column segmentations and reviewed before final severity assignment; formal interobserver reliability for the final categories was not assessed. A convolutional neural network classifier was developed using a pretrained ResNet-50 backbone from RadImageNet, predicting three output classes. To enhance generalizability, data augmentation techniques - including horizontal flipping, random brightness and contrast adjustments, and resizing to 224 × 224 pixels - were applied for 100 epochs. Model performance was evaluated on the test set using metrics such as accuracy, precision, recall, F1-score, Cohen’s Kappa, Matthews Correlation Coefficient, and area under the curve receiver operating characteristic (AUROC). Approximate 95% confidence intervals were calculated for key test-set metrics. Smooth Grad-CAM + + was employed to generate class activation maps for visual interpretability. Results The fine-tuned RadImageNet model achieved an overall accuracy of 76.2% (95% CI, 69.2%-82.1%) and AUROC scores of 0.853 (95% CI, 0.767–0.939) for no scoliosis, 0.766 (95% CI, 0.693–0.839) for mild scoliosis, and 0.843 (95% CI, 0.723–0.962) for severe scoliosis, indicating acceptable but variable class discrimination. The model demonstrated its highest performance for mild scoliosis with a precision of 0.79 (95% CI, 0.72–0.85), recall of 0.90 (95% CI, 0.84–0.95), and an F1 score of 0.85 (95% CI, 0.79–0.89). Misclassifications predominantly occurred between adjacent severity levels. Additionally, class activation maps confirmed that the thoracic and lumbar regions were key contributors to the model’s predictions. Conclusions Fine-tuning a pretrained medical image CNN, specifically a RadImageNet-based ResNet-50 model, for scoliosis grading may support adjunctive scoliosis severity categorization; however, it does not directly measure Cobb angle, progression, sagittal balance, or surgical-planning parameters and requires external validation before clinical screening use.

  • Fragility of randomized controlled trials comparing cemented and cementless components in total knee arthroplasty: A systematic review

    by Jennifer Yu, Avanish Yendluri, Suraj A. Dhanjani, John K. Cordero, John J. Corvi, Michael N. Megafu, Calin S. Moucha, Brett L. Hayden, Robert L. Parisien

  • Youth and adolescent male ice hockey players at increased risk of fractures: a 20-year analysis of national injury data

    by Luca M. Valdivia, Charu Jain, Uma Balachandran, Niklas H. Koehne, Jennifer Yu, Christopher A. White, Salvatore Capotosto, Bradford O. Parsons, David A. Forsh, Robert L. Parisien

    Ice hockey is a high-impact sport that can put athletes in high-risk situations for injuries. While the game has shifted toward speed over physicality, fractures remain a major concern. Therefore, it would be valuable to observe whether this had an effect on fracture rates in men's ice hockey. The purpose of this study was to examine the frequency and trends of fractures sustained in men's ice hockey in the United States. All men's ice hockey-related fractures presented to United States EDs from 1 January 2004 to 31 December 2023 were queried using the National Electronic Injury Surveillance System (NEISS). For each fracture, patient disposition, demographics, and injury site were recorded. National estimates (NE) were calculated using the NEISS statistical sample weight. Injury trends were evaluated by linear regression modeling. A NE of 58,910 ice hockey-related fractures occurred among male players between 2004 and 2023. There was a qualitative decrease in annual fractures (p = 0.208) from 2004 to 2023. There was a significant decrease from 2004 to 2019 (p = 0.001) followed by a 172.6% increase to a record high number of fractures in 2022 (NE = 3,953). The hand was the most commonly fractured part of the body (NE = 17,400, 29.5%). All body regions, except the leg, saw increases in annual injury frequency, including record highs for head, trunk, and shoulder fractures. Adolescents (ages 13-19 years) had the highest fracture rates (56.6%) amongst age groups. Fracture hospitalization rate was 5.5% overall (NE = 3,264). While the overall number of fractures amongst male ice hockey players decreased yearly until the COVID-19 pandemic, a sharp increase in fractures was observed, reaching a record high in 2022. Additionally, the hand was the most commonly fractured part of the body, and adolescents have the highest risk of fractures from ice hockey.

  • Sports-Related Concussions and Closed Head Injuries in Pediatric Female Soccer Athletes: A 10-year Analysis of National Injury Data

    by Jennifer Yu, Avanish Yendluri, Julian V. Javier, Kareem S. Mohamed, Dennis M. Bienstock, Christon Darden, Salvatore Capotosto, Arianna L. Gianakos, Rachel M. Frank, Robert L. Parisien

    Background: Female athletes are underrepresented in sports injury research, particularly regarding sports-related concussions (SRCs) and closed head injuries (CHIs). This study analyzes injury mechanisms in pediatric female soccer players to better address risks in this population. Hypothesis: It was hypothesized that soccer-related SRCs/CHIs in pediatric female players are most commonly caused by ball contact and direct player collisions, with a decrease in incidence over time due to improved safety measures Study Design: Descriptive epidemiological. Level of Evidence: Level 4. Methods: The National Electronic Injury Surveillance System (NEISS) was queried for soccer-related SRCs/CHIs in girls aged 2 to 18 years from 2014 to 2023. Injury narratives were reviewed to identify mechanisms. National estimates (NEs) were calculated using NEISS statistical weights, and linear regression evaluated trends over time. Results: SRCs/CHIs comprised 16.1% of soccer-related injuries in females aged 2 to 18 years, with 84,519 nationally estimated injuries (2946 NEISS cases). Players aged 14 to 18 years accounted for 63.8% of cases. The most common mechanism was being hit by a soccer ball kicked by another player (35.0%). Player contact, including head-to-head collisions and rough play, caused 25.2% of injuries. For goalies, being kicked or kneed in the head was the leading mechanism (34.6%). A significant decrease in injuries was observed over time ( P < 0.05). Conclusion: Female pediatric soccer head injuries decreased significantly from 2014 to 2023, with most injuries caused by ball contact or player collisions. These findings support implementing safety initiatives, such as headgear, concussion education, and gameplay modifications, to reduce injury risks based on the mechanisms identified. Clinical Relevance: This study highlights the need for safety measures that can be applied broadly across age groups to mitigate soccer-related head injuries in pediatric female athletes.

  • Female high school soccer injuries presenting to US emergency departments: A 10-year analysis of national injury data

    by Jennifer Yu, Avanish Yendluri, Julian V. Javier, Sohan S. Shah, Jeffrey Yu, Dennis M. Bienstock, Christon Darden, Salvatore Capotosto, Rachel M. Frank, Theodore J. Ganley, et al.


  • Body mass index does not increase risk of complications after reverse total shoulder arthroplasty

    by Salvatore Capotosto, Christopher White, Jennifer Yu, John J. Corvi, Hulaimatu Jalloh, Paul J. Cagle


  • Sociodemographic variables are rarely included in femoral neck fracture randomized controlled trials: A systematic review

    by Avanish Yendluri, Jennifer Yu, Carolina Stocchi, Mark Kurapatti, Nikan K. Namiri, Junho Song, John J. Corvi, John K. Cordero, Steven Yacovelli, Brett L. Hayden, et al.

  • Automated Scoliosis Cobb Angle Classification in Biplanar Radiograph Imaging With Explainable Machine Learning Models

    by Jennifer Bs Yu, Yash Mse Lahoti, Kyle C. BS McCandless, Nikan K. Namiri, Matthew S. Miyasaka, Hamza Mcs Ahmed, Junho Song, John J. Corvi, Daniel C. Berman, Samuel K. Cho, et al.

    Study Design.: Retrospective cohort study. Objective.: To quantify the pathology of the spine in patients with scoliosis through one-dimensional feature analysis. Summary of Background Data.: Biplanar radiograph (EOS) imaging is a low-dose technology offering high-resolution spinal curvature measurement, crucial for assessing scoliosis severity and guiding treatment decisions. Machine learning (ML) algorithms, utilizing one-dimensional image features, can enable automated Cobb angle classification, improving accuracy and efficiency in scoliosis evaluation while reducing the need for manual measurements, thus supporting clinical decision-making. Methods.: This study used 816 annotated AP EOS spinal images with a spine segmentation mask and a 10° polynomial to represent curvature. Engineered features included the first and second derivatives, Fourier transform, and curve energy, normalized for robustness. XGBoost selected the top 32 features. The models classified scoliosis into multiple groups based on curvature degree, measured through Cobb angle. To address the class imbalance, stratified sampling, undersampling, and oversampling techniques were used, with 10-fold stratified K-fold cross-validation for generalization. An automatic grid search was used for hyperparameter optimization, with K-fold cross-validation (K=3). Results.: The top-performing model was Random Forest, achieving an ROC AUC of 91.8%. An accuracy of 86.1%, precision of 86.0%, recall of 86.0%, and an F1 score of 85.1% were also achieved. Of the three techniques used to address class imbalance, stratified sampling produced the best out-of-sample results. SHAP values were generated for the top 20 features, including spine curve length and linear regression error, with the most predictive features ranked at the top, enhancing model explainability. Conclusions.: Feature engineering with classical ML methods offers an effective approach for classifying scoliosis severity based on Cobb angle ranges. The high interpretability of features in representing spinal pathology, along with the ease of use of classical ML techniques, makes this an attractive solution for developing automated tools to manage complex spinal measurements.


  • Top Cited Papers
    Open Access

    Increasing Incidence of Pickleball Injuries Presenting to US Emergency Departments: A 10-year Epidemiologic Analysis of Mechanisms and Trends

    by Jennifer Yu, Avanish Yendluri, Gabriel S. Linden, Nikan K. Namiri, John J. Corvi, Junho Song, Bradford O. Parsons, Robert L. Parisien

    Background: Pickleball is one of the fastest-growing sports in the United States. It is popular among seniors but has recently grown across all age groups. As pickleball has gained interest, its corresponding injury burden has also increased. Purpose/Hypothesis: The purpose of this study was to identify pickleball-related injury trends and underlying mechanisms across different age groups presenting to US emergency departments. It was hypothesized that (1) pickleball-related injuries would increase significantly between 2013 and 2022, (2) older age groups would have the greatest frequency of injury, and (3) falls would be the most common mechanism of injury. Study Design: Descriptive epidemiology study. Methods: The National Electronic Injury Surveillance System (NEISS) was queried for pickleball injuries between January 1, 2013, and December 31, 2022. The demographic information, injury site, and diagnosis were recorded, and case narratives were reviewed to identify the injury mechanism. The statistical sample weight assigned for NEISS cases by hospital was used to calculate national estimates (NEs). Injury trends over time were evaluated by linear regression. Results: A total of 1110 NEISS cases representing 66,350 nationally estimated pickleball-related injuries were included. The mean age was 64 ± 14.7 years (range, 4-93 years). Most injuries occurred in those aged 65 to 80 years (NEISS cases = 643; NE = 40,507 [61.1%]), followed by those aged 35 to 64 years (NEISS cases = 349; NE = 20,785 [31.3%]). Falls were the most common injury mechanism overall (NEISS cases = 699; NE = 43,434 [65.5%]). However, being hit with a paddle and a twist/inversion were the most common mechanisms for ages 0 to 17 and 18 to 34 years, respectively. Fractures were the most common diagnosis (NEISS cases = 363; NE = 21,703 [32.7%]), followed by strains/sprains (NEISS cases = 327; NE = 20,419 [30.8%]). Also, 69.1% of all fractures occurred in female players. The wrist was the most injured, accounting for 12.7% of all injuries. Significant increases in pickleball injuries were seen over time ( P < .001) and across all age groups ( P < .05). Conclusion: There was a significant increase in pickleball-related injuries presenting between 2013 and 2022, with most injuries arising from a fall. Notably, injuries decreased by 14% between 2019 and 2020 but rose again by 41% between 2020 and 2021 during the coronavirus 2019 pandemic. Older players disproportionately presented to the emergency department with pickleball-related injuries. Increased awareness and comprehensive injury prevention initiatives may be warranted to address the considerable prevalence of fractures and overuse/strain injuries.