Scalable Injury-Risk Screening in Baseball Pitching From Broadcast Video
This work provides a scalable and accessible method for injury-risk screening in baseball pitching, enabling broader application beyond professional venues where expensive multi-camera systems are unavailable.
This paper developed a monocular video pipeline that extracts 18 biomechanical metrics from broadcast baseball footage, achieving sub-degree agreement (MAE < 1°) for 16/18 metrics compared to gold-standard measurements on 13 professional pitchers. An injury screening model built on these metrics achieved an AUC of 0.811 for Tommy John surgery and 0.825 for significant arm injuries across 7,348 pitchers.
Injury prediction in pitching depends on precise biomechanical signals, yet gold-standard measurements come from expensive, stadium-installed multi-camera systems that are unavailable outside professional venues. We present a monocular video pipeline that recovers 18 clinically relevant biomechanics metrics from broadcast footage, positioning pose-derived kinematics as a scalable source for injury-risk modeling. Built on DreamPose3D, our approach introduces a drift-controlled global lifting module that recovers pelvis trajectory via velocity-based parameterization and sliding-window inference, lifting pelvis-rooted poses into global space. To address motion blur, compression artifacts, and extreme pitching poses, we incorporate a kinematics refinement pipeline with bone-length constraints, joint-limited inverse kinematics, smoothing, and symmetry constraints to ensure temporally stable and physically plausible kinematics. On 13 professional pitchers (156 paired pitches), 16/18 metrics achieve sub-degree agreement (MAE $< 1^{\circ}$). Using these metrics for injury prediction, an automated screening model achieves AUC 0.811 for Tommy John surgery and 0.825 for significant arm injuries on 7,348 pitchers. The resulting pose-derived metrics support scalable injury-risk screening, establishing monocular broadcast video as a viable alternative to stadium-scale motion capture for biomechanics.