RESUMO
Individuals in isolated and extreme environments can experience debilitating side-effects including significant decreases in fat-free mass (FFM) from disuse and inadequate nutrition. The objective of this study was to determine the strengths and weaknesses of three-dimensional optical (3DO) imaging for monitoring body composition in either simulated or actual remote environments. Thirty healthy adults (ASTRO, male = 15) and twenty-two Antarctic Expeditioners (ABCS, male = 18) were assessed for body composition. ASTRO participants completed duplicate 3DO scans while standing and inverted by gravity boots plus a single dual-energy X-ray absorptiometry (DXA) scan. The inverted scans were an analog for fluid redistribution from gravity changes. An existing body composition model was used to estimate fat mass (FM) and FFM from 3DO meshes. 3DO body composition estimates were compared to DXA with linear regression and reported with the coefficient of determination (R2) and root mean square error (RMSE). ABCS participants received only duplicate 3DO scans on a monthly basis. Standing ASTRO meshes achieved an R2 of 0.76 and 0.97 with an RMSE of 2.62 and 2.04 kg for FM and FFM, while inverted meshes achieved an R2 of 0.52 and 0.93 with an RMSE of 2.84 and 3.23 kg for FM and FFM, respectively, compared to DXA. For the ABCS arm, mean weight, FM, and FFM changes were -0.47, 0.06, and -0.54 kg, respectively. Simulated fluid redistribution decreased the accuracy of estimated body composition values from 3DO scans. However, FFM stayed robust. 3DO imaging showed good absolute accuracy for body composition assessment in isolated and remote environments.
RESUMO
BACKGROUND: Body shape, an intuitive health indicator, is deterministically driven by body composition. We developed and validated a deep learning model that generates accurate dual-energy X-ray absorptiometry (DXA) scans from three-dimensional optical body scans (3DO), enabling compositional analysis of the whole body and specified subregions. Previous works on generative medical imaging models lack quantitative validation and only report quality metrics. METHODS: Our model was self-supervised pretrained on two large clinical DXA datasets and fine-tuned using the Shape Up! Adults study dataset. Model-predicted scans from a holdout test set were evaluated using clinical commercial DXA software for compositional accuracy. RESULTS: Predicted DXA scans achieve R2 of 0.73, 0.89, and 0.99 and RMSEs of 5.32, 6.56, and 4.15 kg for total fat mass (FM), fat-free mass (FFM), and total mass, respectively. Custom subregion analysis results in R2s of 0.70-0.89 for left and right thigh composition. We demonstrate the ability of models to produce quantitatively accurate visualizations of soft tissue and bone, confirming a strong relationship between body shape and composition. CONCLUSIONS: This work highlights the potential of generative models in medical imaging and reinforces the importance of quantitative validation for assessing their clinical utility.
Body composition, measured quantities of muscle, fat, and bone, is typically assessed through dual energy X-ray absorptiometry (DXA) scans, which requires specialized equipment, trained technicians and involves exposure to radiation. Exterior body shape is dependent on body composition and recent technological advances have made three-dimensional (3D) scanning for body shape accessible and virtually ubiquitous. We developed a model which uses 3D body surface scan inputs to generate DXA scans. When analyzed with commercial software that is used clinically, our model generated images yielded accurate quantities of fat, lean, and bone. Our work highlights the strong relationship between exterior body shape and interior composition. Moreover, it suggests that with enhanced accuracy, such medical imaging models could be more widely adopted in clinical care, making the analysis of body composition more accessible and easier to obtain.