Fig. 1: Schematic of study design. | Nature Communications

Fig. 1: Schematic of study design.

From: Genetic Insights into Head-to-Body Ratios Via Deep Learning-Based Image Segmentation and Implications for Common Diseases

Fig. 1: Schematic of study design.The alternative text for this image may have been generated using AI.

a We constructed multiple deep learning models to perform quality control and processing on the large-scale DXA image data from the UK Biobank. We constructed a series of image segmentation models, extracted head and body dimensions data, and calculated 10 types of head-to-body ratio (HBR). b Combining the individual genotype data from UKB, we conducted GWASs to identify genetic loci associated with HBRs. Using various analytical methods, we obtained an optimized set of HBR genes and performed functional annotation, MGI phenotypes annotation, and single-cell disease relevance score. c We analyzed the role of HBRs in human evolution. d We explored the association and the causality, between HBRs and common diseases. KYA: kilo years ago, MYA: million years ago, p.c.w: post-conceptional week. Panels c, d were partly generated using Servier Medical Art, provided by Servier, licensed under the Creative Commons Attribution 4.0 International License. All organism silhouettes are from PhyloPic. DXA images are reproduced by kind permission of UK Biobank ©.

Back to article page