Estimating Visceral Adiposity from Wrist-Worn Accelerometry

2025-06-10

AI Summary

Visceral adipose tissue (VAT) quantification—critical for metabolic risk assessment—is typically limited by the cost and inaccessibility of imaging modalities. Method: This study proposes a noninvasive, wrist-worn accelerometer–based VAT estimation method leveraging gait and sleep-derived engineering features. We introduce a dual-path temporal modeling framework: (1) a day-level Transformer to capture long-term activity patterns, and (2) a fine-grained path using 10-second frame embeddings aggregated across multiple days. The model jointly integrates ridge regression and deep learning, incorporating demographic covariates for calibration. Contribution/Results: Evaluated on the large-scale NHANES cohort, our approach achieves a prediction correlation coefficient of r = 0.86 for VAT—substantially outperforming conventional anthropometric proxies (e.g., BMI, waist circumference). Results demonstrate that wrist-accelerometry signals encode robust metabolic risk information, establishing a novel paradigm for wearable-enabled, precision metabolic health assessment.

Technology Category

• Computer Vision: Biometrics, Face, Gesture & Pose
• Intelligent Robots: State Estimation
• Humans and AI: Voting

Application Category

• User Modeling, Personalization and Recommendation: Metrics for user behavior and evaluating success
• Security and Privacy: Large-scale security measurements
• Search and Retrieval-Augmented AI: Web evaluation methodologies and metrics

Abstract

Visceral adipose tissue (VAT) is a key marker of both metabolic health and habitual physical activity (PA). Excess VAT is highly correlated with type 2 diabetes and insulin resistance. The mechanistic basis for this pathophysiology relates to overloading the liver with fatty acids. VAT is also a highly labile fat depot, with increased turnover stimulated by catecholamines during exercise. VAT can be measured with sophisticated imaging technologies, but can also be inferred directly from PA. We tested this relationship using National Health and Nutrition Examination Survey (NHANES) data from 2011-2014, for individuals aged 20-60 years with 7 days of accelerometry data (n=2,456 men; 2,427 women) [1]. Two approaches were used for estimating VAT from activity. The first used engineered features based on movements during gait and sleep, and then ridge regression to map summary statistics of these features into a VAT estimate. The second approach used deep neural networks trained on 24 hours of continuous accelerometry. A foundation model first mapped each 10s frame into a high-dimensional feature vector. A transformer model then mapped each day's feature vector time series into a VAT estimate, which were averaged over multiple days. For both approaches, the most accurate estimates were obtained with the addition of covariate information about subject demographics and body measurements. The best performance was obtained by combining the two approaches, resulting in VAT estimates with correlations of r=0.86. These findings demonstrate a strong relationship between PA and VAT and, by extension, between PA and metabolic health risks.

Problem

• Estimating visceral fat from wrist-worn accelerometer data
• Linking physical activity to metabolic health risks
• Improving VAT measurement accuracy using AI and covariates

Innovation

• Estimating VAT using wrist-worn accelerometry data
• Combining ridge regression and deep neural networks
• Incorporating demographics for improved VAT accuracy