Concurrent Assessment and Inter-Session Repeatability of Markerless Motion Capture
Principal Investigator: Hector Carbajal Mendez, MS
Advances in computer vision and deep learning are transforming how human movement is measured. This project evaluated whether markerless motion capture systems can achieve the accuracy and reliability of traditional marker-based methods.
Conducted at the UCSF Human Performance Center, the study involved 10 healthy participants performing sit-to-stand and walking tasks while data were collected simultaneously using both systems. Results showed that markerless technology achieved comparable accuracy in sagittal and frontal plane joint-angle measurements, with improved consistency across sessions and reduced variability.
Additional analyses explored the effects of video resolution and clothing on measurement accuracy. Overall, the findings support markerless motion capture as a practical, noninvasive alternative for biomechanical analysis in controlled environments.
Findings were presented at the 2024 American Society of Biomechanics (ASB) Annual Conference.
Turning Everyday Wearable Data into Baseline Performance Scores for Recovery Tracking
Principal Investigator: Meir Marmor, MD
Can everyday wearable devices help track recovery after injury? This study explored whether data collected from devices such as iPhones and Apple Watches can predict clinical performance metrics.
Eleven participants provided one year of routine activity data, including step count, stair count, heart rate, respiratory rate, and heart rate variability, which was compared with laboratory-based performance measures such as VO₂ max, vertical jump, and mobility tests.
Strong correlations were identified between wearable data and peak performance outcomes. Machine learning models further demonstrated the ability to estimate fitness levels with moderate accuracy (R² ≈ 0.6).
These proof-of-concept findings suggest that commonly collected wearable data may offer a convenient, scalable method for estimating baseline or pre-injury performance, potentially enhancing recovery tracking and clinical decision-making.
Findings were presented at the 2025 Western Orthopaedic Association Annual Meeting.
Biomechanical Analysis of Foot Health in Diabetic Neuropathy and Chronic Kidney Disease
Principal Investigator: Victor Cheuy, PhD
This study investigates how chronic kidney disease (CKD) and diabetic neuropathy together increase the risk of foot complications and amputation.
Using advanced tools at the UCSF Human Performance Center—including motion capture for multisegment foot analysis, physiological aging assessments, and gait pedography—researchers incorporated high-resolution CT imaging to identify key differences in patients with both conditions. Compared with controls and individuals with neuropathy alone, these patients demonstrated reduced bone quality, increased plantar pressure, and impaired ankle function.
These early biomechanical changes may contribute to deformities, ulcers, and ultimately, amputation.
This work led to a multi-site NIH R01 study, now underway in collaboration with Washington University in St. Louis, High Point University, and Wake Forest University. The ongoing research aims to better understand risk factors and improve prevention strategies for patients with diabetes and chronic kidney disease, with a particular focus on reducing amputation risk.