Bio


I work at the intersection of machine learning, medical imaging, and biomechanics. I'm committed to developing tools that bridge gaps between computational methods, musculoskeletal research, and clinical care - enabling more precise analyses, efficient workflows, and improved surgical decision-making.

Stanford Advisors


All Publications


  • BMI and Varus Malalignment Compound to Define a High-Risk Phenotype for Compartment-Specific Knee Osteoarthritis Progression. Osteoarthritis and cartilage White, M. S., Kogan, F., Delp, S. L., Chu, C. R., Sherman, S. L., Pai S, A., Gold, G. E., Chaudhari, A. S., Gatti, A. A. 2026

    Abstract

    Knee osteoarthritis (KOA) is a leading cause of disability, yet which patients will experience structural decline remains unclear. Body mass index (BMI) and lower limb alignment are established risk factors for KOA, but their independent and interactive effects on compartment-specific cartilage loss and total knee replacement (TKR) have not been characterized at scale.We analyzed 5,930 limbs from 3,191 participants in the Osteoarthritis Initiative followed over 0-60 months. Cartilage thickness in the weight-bearing medial and lateral femur and tibia was quantified, and lower limb alignment was measured using hip-knee-ankle (HKA) angle obtained from full-limb radiographs. Linear mixed-effects models estimated the associations of BMI and lower limb alignment with longitudinal changes in cartilage thickness, modeled as interactions with time and with each other and adjusted for age, sex, and baseline Kellgren-Lawrence grade. Logistic regression with participant-clustered standard errors modeled TKR odds within five and ten years after each limb's baseline.In the medial compartment, BMI and varus alignment interacted, with their combined effect exceeding the sum of independent effects (femur: β = 0.0016 mm/year per 1-SD BMI x 1-SD HKA, 95% CI [0.0001, 0.0031]; tibia: β = 0.0018, 95% CI [0.0008, 0.0028]). In the lateral compartment, alignment, but not BMI, was associated with faster cartilage loss, and its association was nonlinear (time x HKA2: femur β = -0.0032 mm/year, 95% CI [-0.0038, -0.0027]; tibia β = -0.0033, 95% CI [-0.0040, -0.0026]). TKR odds increased 1.11-fold at five years (95% CI [1.04, 1.18]) and 1.14-fold at ten years (95% CI [1.08, 1.21]) per 1-degree greater absolute HKA deviation from the radiographically healthy reference angle (-1.0°). BMI did not modify the association between HKA deviation and TKR.BMI and lower limb alignment influence structural KOA progression through compartment-specific pathways. The compounding interaction in the medial compartment identifies high BMI combined with varus malalignment as a high-risk phenotype for medial structural progression.

    View details for DOI 10.1016/j.joca.2026.11.003

    View details for PubMedID 43020420

  • Neural shape models enable cohort-scale simulation of knee cartilage mechanics. medRxiv : the preprint server for health sciences Gatti, A. A., Marusich, K. R., Clouthier, A., White, M. S., Ong, C., Esrafilian, A., Smith, C., Kolesar, J., Tan, T., Uhlrich, S., Chu, C., Sherman, S., Gold, G. E., Delp, S. L., Kogan, F., Chaudhari, A. S. 2026

    Abstract

    Knee osteoarthritis affects over 350 million people worldwide, is a leading cause of disability, and carries substantial costs. Cartilage mechanical load is believed to play a critical role in osteoarthritis pathophysiology, yet clinical biomechanics research uses external measures of joint loading that do not capture tissue-level load. We developed a fully-automated pipeline that builds biomechanical knee models from a single MRI using deep learning-based segmentation and neural shape models, and elastic foundation contact to simulate gait. Applied to 756 knees from the Osteoarthritis Initiative, each simulated under the same 10 standardized gaits, the pipeline converged in 99.6% of 7,560 simulations. Medial tibial pressure decreased from healthy to incident to mild osteoarthritis as disease-related shape changes enlarged contact area, an effect reproduced in simulations of synthetic knees. Yet within knees, medial cartilage thinned fastest over six years in regions of highest pressure, with stronger relationships in incident and mild osteoarthritis knees. Automated simulations provided cohort-scale, tissue-level support for mechanical theories of osteoarthritis progression and a framework for prospectively quantifying cartilage load to guide mechanical interventions.

    View details for DOI 10.64898/2026.10.29.26366866

    View details for PubMedID 42990758

    View details for PubMedCentralID PMC13734216

  • Neural Shape Modeling Reveals Early and Progressive Femoral Bone Shape and Cartilage Thickness Changes After Anterior Cruciate Ligament Reconstruction. Journal of orthopaedic research : official publication of the Orthopaedic Research Society Pai S, A., White, M. S., Black, M. S., Young, K. A., Sherman, S. L., Chu, C. R., Williams, A., Gold, G. E., Kogan, F., Hargreaves, B. A., Chaudhari, A. S., Gatti, A. A. 2026; 44 (9): e70274

    Abstract

    Early and longitudinal shape changes in femoral bone and cartilage after anterior cruciate ligament reconstruction (ACLR) remains poorly understood. It is also unknown whether post traumatic shape changes follow similar trajectory as progressive alterations seen in idiopathic osteoarthritis (OA). Two neural shape models (NSMs)-B-NSM (bone-only) and BC-NSM (bone+cartilage) encoded femoral bone and cartilage shapes from 17 ACLR, contralateral, and healthy control knees at 3 weeks and 3, 9, 18, and 30 months post-ACLR. Knee OA severity shape scores (B-Score, BC-Score), bone surface area, and cartilage thickness were computed. Linear mixed-effect models assessed cross-sectional and longitudinal changes in these metrics; etap 2 evaluated sensitivity to change. Cosine similarity quantified post-ACLR shape change trajectories compared to idiopathic OA progression. At 3 weeks, ACLR knee B-Scores were significantly lower (p=0.010) than contralateral knees from surgical notchplasty. Over 30 months, ACLR knee B- and BC-Scores significantly increased compared to baseline (p<0.001 for both). Bone surface area (etap 2=0.01) increased from periarticular osteophyte lipping with concomitant region-specific cartilage thickening and thinning (etap 2=0.02). B-Score (etap 2=0.45) was more sensitive to longitudinal change than BC-Score (etap 2=0.30). ACLR bone shape changes aligned with idiopathic-OA progression (cosine similarity=0.22), while cartilage did not (cosine similarity=0.08). Early and progressive joint remodeling were observed post-ACLR. Both B- and BC-Score were more sensitive than bone surface area and cartilage thickness measures in quantifying longitudinal changes in ACLR knees. While ACLR bone shape features exhibited a linear progressive pattern similar to idiopathic OA, cartilage thickness changes were nonlinear. These findings support the use of NSM-based metrics for tracking post-surgical joint remodeling in future clinical trials.

    View details for DOI 10.1002/jor.70274

    View details for PubMedID 42754948

  • Toward measurement of muscle forces during movement. Science robotics Tan, T., White, M. S., Delp, S. L. 2026; 11 (116): eaej8562

    Abstract

    Ultra-wideband radar identifies components of a muscle's contractile state to estimate muscle force.

    View details for DOI 10.1126/scirobotics.aej8562

    View details for PubMedID 42525727

  • Dynamic Fibro-adipogenic Progenitor States Contribute to Loss of Muscle Quality After Injury in Humans Gonzalez-Velez, S., Keeble, A., Owen, A., Wohlgemuth, R., Thomas, N., Dobis, M., Thompson, H., White, M., Johnson, D., Stone, A., Wen, Y., Noehren, B., Fry, C. AMER PHYSIOLOGICAL SOC. 2026
  • BMI and Varus Malalignment Compound to Define a High-Risk Phenotype for Compartment-Specific Knee Osteoarthritis Progression. medRxiv : the preprint server for health sciences White, M. S., Kogan, F., Delp, S. L., Chu, C. R., Sherman, S. L., S, A. P., Gold, G. E., Chaudhari, A. S., Gatti, A. A. 2026

    Abstract

    Knee osteoarthritis (KOA) is a leading cause of disability, yet which patients will experience structural decline remains unclear. Body mass index (BMI) and lower limb alignment are established risk factors for KOA, but their independent and interactive effects on compartment-specific cartilage loss and total knee replacement (TKR) have not been characterized at scale.We analyzed 5,832 limbs from 3,016 participants in the Osteoarthritis Initiative followed over 7 years. Cartilage thickness in the weight-bearing medial and lateral femur and tibia was quantified, and lower limb alignment was measured using hip-knee-ankle (HKA) angle obtained from full-limb radiographs. Linear mixed-effects models estimated the independent and interactive effects of BMI and lower limb alignment on longitudinal cartilage thinning, and mixed-effects logistic regression modeled TKR risk.In the medial compartment, BMI and varus alignment interacted multiplicatively, with their combined effect exceeding the sum of independent contributions (femur: p = 0.011; tibia: p< 0.001). At +10kg/m2 BMI and +10° varus, the rate of medial femur cartilage thinning was 243.5% faster than the reference rate. In the lateral compartment, BMI and valgus alignment were independently associated with faster cartilage thinning, with no significant interaction. TKR risk increased exponentially with HKA deviation (odds ratio [OR] = 1.38 per 1°; ~five-fold at 5° malalignment) but was not associated with BMI.BMI and lower limb alignment influence structural KOA progression through compartment-specific pathways. The multiplicative interaction in the medial compartment identifies high BMI combined with varus malalignment as a discrete high-risk phenotype, with implications for clinical risk stratification and disease-modifying intervention design.

    View details for DOI 10.64898/2026.04.15.26350819

    View details for PubMedID 42064958

    View details for PubMedCentralID PMC13127543

  • NEURAL SHAPE MODEL-BASED SCORE OUTPERFORMS CONVENTIONAL MEASURES IN QUANTIFYING EARLY AND PROGRESSIVE JOINT REMODELING IN ANTERIOR CRUCIATE LIGAMENT RECONSTRUCTED KNEES Pai, S., White, M. S., Black, M. S., Young, K., Sherman, S., Chu, C., Williams, A., Gold, G., Kogan, F., Hargreaves, B., Chaudhari, A., Gatti, A. ELSEVIER SCI LTD. 2026: S271-S272
  • SEX-SPECIFIC ALTERATIONS IN THIGH MUSCLE QUALITY IN INDIVIDUALS WITH KNEE OSTEOARTHRITIS Belibi, F., Vainberg, Y., White, M. S., Gatti, A., Weber, K. A., Gold, G., Kogan, F., Goyal, A. ELSEVIER SCI LTD. 2026: S262-S263
  • VARUS ALIGNMENT AND BMI JOINTLY ACCELERATE MEDIAL FEMORAL CARTILAGE THINNING: DATA FROM THE OSTEOARTHRITIS INITIATIVE White, M. S., Kogan, F., Delp, S. L., Gold, G., Chaudhari, A., Gatti, A. ELSEVIER SCI LTD. 2026: S67-S68
  • Osteoarthritis year in review 2025: Imaging. Osteoarthritis and cartilage Harkey, M. S., Gatti, A. A., Jansen, M. P., White, M. S., Tolzman, J., Parmar, A., Frederick, I., Link, T. M., Hart, H. F. 2025

    Abstract

    To conduct a narrative review of selected imaging studies published in the past year, with a focus on how emerging methods and multi-tissue assessments are contributing to disease characterization, early detection, and intervention strategies.We conducted a comprehensive search strategy of PubMed, Embase, and CINAHL for original English-language human studies on imaging in osteoarthritis, published between March 2024 and March 2025. From the 262 full-text studies that met inclusion criteria, we identified a set of representative studies through independent review and group discussion. These studies are organized into five themes that emerged from this study: effusion-synovitis, infrapatellar fat pad, muscle, subchondral bone, and subcutaneous fat. Within each theme, one highlighted study and several additional studies were synthesized, followed by a summary of cross-cutting insights and future directions.A total of 26 studies were included. Collectively, the studies reinforce the importance of viewing osteoarthritis as a condition involving multiple joint structures, with inflammation as a recurring feature across tissues. Muscle quality and local adiposity emerged as consistent imaging biomarkers, while accessible tools such as ultrasound and AI-enhanced radiography demonstrated potential for scalable clinical application. Links between imaging findings and biomechanics, particularly gait and loading patterns, also featured prominently. Cross-cutting themes emphasized the value of considering multiple tissues, expanding the reach of imaging to clinical and community settings, and integrating imaging with functional measures.Imaging research in the osteoarthritis continues to broaden in scope and application. The studies highlighted in this review illustrate how advances in methods and metrics are improving opportunities for earlier detection, more precise assessment of symptoms, and the development of targeted, structure-informed interventions.

    View details for DOI 10.1016/j.joca.2025.11.004

    View details for PubMedID 41241372

  • Open-Source AI for Vastus Lateralis and Adipose Tissue Segmentation to Assess Muscle Size and Quality ULTRASOUND IN MEDICINE AND BIOLOGY White, M. S., Horikawa-Strakovsky, A., Mayer, K. P., Noehren, B. W., Wen, Y. 2025; 51 (12): 2276-2280

    Abstract

    Ultrasound imaging is a clinically feasible method for assessing muscle size and quality, but manual processing is time-consuming and difficult to scale. Existing artificial intelligence (AI) models measure muscle cross-sectional area, but they do not include assessments of muscle quality or account for the influence of subcutaneous adipose tissue thickness on echo intensity measurements. We developed an open-source AI model to accurately segment the vastus lateralis and subcutaneous adipose tissue in B-mode images for automating measurements of muscle size and quality.The model was trained on 612 ultrasound images from 44 participants who had anterior cruciate ligament reconstruction. Model generalizability was evaluated on a test set of 50 images from 14 unique participants. A U-Net architecture with ResNet50 backbone was used for segmentation. Performance was assessed using the Dice coefficient and Intersection over Union (IoU). Agreement between model predictions and manual measurements was evaluated using intraclass correlation coefficients (ICCs), R² values and standard errors of measurement (SEM).Dice coefficients were 0.9095 and 0.9654 for subcutaneous adipose tissue and vastus lateralis segmentation, respectively. Excellent agreement was observed between model predictions and manual measurements for cross-sectional area (ICC = 0.986), echo intensity (ICC = 0.991) and subcutaneous adipose tissue thickness (ICC = 0.996). The model demonstrated high reliability with low SEM values for clinical measurements (cross-sectional area: 1.15 cm², echo intensity: 1.28-1.78 a.u.).We developed an open-source AI model that accurately segments the vastus lateralis and subcutaneous adipose tissue in B-mode ultrasound images, enabling automated measurements of muscle size and quality.

    View details for DOI 10.1016/j.ultrasmedbio.2025.08.008

    View details for Web of Science ID 001598156500003

    View details for PubMedID 40947375

    View details for PubMedCentralID PMC12453877

  • Associations Between Body Mass Index, Gait Biomechanics, and In Vivo Cartilage Function After Exercise in Those With Anterior Cruciate Ligament Reconstruction AMERICAN JOURNAL OF SPORTS MEDICINE Garcia, S. A., White, M. S., Gallegos, J., Balza, I., Kahan, S., Palmieri-Smith, R. M. 2024; 52 (13): 3295-3305

    Abstract

    Both high body mass index (BMI) and anterior cruciate ligament reconstruction (ACLR) independently influence knee osteoarthritis risk. Preliminary evidence shows the combination of these risk factors leads to poorer recovery and altered biomechanical outcomes after ACLR, but few studies have directly evaluated early changes in cartilage health between normal-BMI and high-BMI groups in this population.To evaluate ultrasound-based measures of cartilage strain and compositional changes (via echo-intensity [EI]) in response to an incline walking stress test between normal-BMI and high-BMI individuals with ACLR. A secondary evaluation was conducted of associations between habitual walking biomechanics (ie, ground-reaction forces, sagittal knee kinetics and kinematics) and cartilage strain and EI outcomes.Controlled laboratory study.Gait biomechanics and femoral trochlear ultrasound analyses were evaluated in 64 participants with ACLR who had normal BMI (BMI < 27.0; n = 40) and high BMI (BMI ≥ 27.0; n = 24). Ultrasound images were collected bilaterally before and after an incline treadmill walk, and medial and lateral trochlear strain and EI changes pre-post exercise were used to compare BMI groups and limbs. Gait outcomes included ground-reaction forces, peak sagittal plane knee moments, angles, and excursions and were used to determine associations with cartilage outcomes in the entire cohort.High-BMI individuals with ACLR exhibited greater medial trochlear cartilage strain in the ACLR limb compared with normal-BMI individuals (approximately 6%; P < .01). In those with high BMI, the ACLR limb exhibited greater medial trochlear strain relative to non-ACLR limbs (approximately 4%; P < .05), but between-limb differences were not observed in the normal-BMI group (P > .05). Medial trochlear EI changes were greater bilaterally in those with high BMI compared with normal-BMI ACLR counterparts (approximately 10%; P < .01). Last, individuals who walked with greater peak knee flexion angles exhibited less medial cartilage strain (ΔR2 = 0.06; P = .025).The data suggested that high BMI affects cartilage functional properties after ACLR, whereas smaller knee flexion angles were associated with larger medial cartilage strain.High-BMI individuals with ACLR may represent a subset of patients exhibiting earlier declines in cartilage functional integrity in response to loading, necessitating additional or more targeted interventions to mitigate disease development.

    View details for DOI 10.1177/03635465241281333

    View details for Web of Science ID 001350124200023

    View details for PubMedID 39503724

  • Associations between Body Mass Index, Gait Mechanics and Trochlear Cartilage Thickness in Those with ACL Reconstruction MEDICINE & SCIENCE IN SPORTS & EXERCISE Garcia, S. A., White, M. S., Gallegos, J., Balza, I., Kahan, S., Palmieri-Smith, R. M. 2024; 56 (9): 1805-1815

    Abstract

    High body mass index (BMI) is a strong predictor of posttraumatic osteoarthritis (OA) after anterior cruciate ligament reconstruction (ACLR). Altered gait mechanics are independently affected by BMI and ACLR, and influence OA risk. Yet, evidence directly assessing the impact of high BMI on gait or cartilage characteristics after ACLR are limited. Here, we evaluated if high BMI moderates associations between gait and trochlear cartilage structure in individuals with ACLR.Treadmill walking biomechanics were evaluated in 40 normal BMI and 24 high BMI participants with ACLR at self-selected speeds. Normalized and absolute peak and cumulative loads (i.e., impulse) were extracted for peak knee flexion and adduction moments (KFM, KAM) and vertical ground reaction force (GRF). Medial and lateral femoral cartilage thickness and medial/lateral thickness ratios were assessed via ultrasound.Those with ACLR and high BMI walked with reduced normalized peak vertical GRFs, and greater absolute peak and cumulative loads compared with normal BMI individuals with ACLR. Those with ACLR and high BMI also exhibited thinner cartilage and greater medial/lateral ratios in ACLR limbs compared with contralateral limbs whereas normal BMI individuals with ACLR exhibited thicker ACLR limb cartilage. Lastly, greater peak KAM and KAM cumulative load were associated with thicker lateral cartilage and lesser medial/lateral thickness ratios, but only in the high BMI group.We observed those with high BMI after ACLR exhibited trochlear cartilage structural alterations not observed in normal BMI patients, whereas differential associations between loading outcomes and cartilage thickness in ACLR knees were observed between groups. Those with high BMI after ACLR may require different therapeutic strategies to optimize joint health in this subset of patients.

    View details for DOI 10.1249/MSS.0000000000003446

    View details for Web of Science ID 001291538300006

    View details for PubMedID 38600643

    View details for PubMedCentralID PMC11326996