Scholarlink

Obesity and overweight have become a major health concern, currently about 30% of the population suffers from obesity and this percentage is expected to increase to 50% by 2035. Obesity accounts for more than five million deaths since 2019 and its costs are about 2.19% of global gross domestic product (GDP). This paper focuses on how the utilization of Variational Autoencoders (VAEs) and the interpretation of the latent space can help in the predictions of the evolution of the patients using clinical data and 3D shape representation of the body. This research aims to bring treatment closer to patients and improve outcomes in the treatment of obesity and overweight without the need of expensive methods. The results show that the proposed method accurately reconstructs 3D models of the human body and clinical data, and the analysis of the latent space shows a new way to study a patient’s progress.

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

A Deep Learning Analysis of 3D Body Shapes and Clinical Data for Understanding Dietetic Interventions

  • 会议论文

  • 发表日期：

    2026年

查看全文
  • Nahuel Garcia-D’Urso,
  • Jorge Azorin-Lopez,
  • Andres Fuster-Guillo,
  • Mario Jerez-Tallon

摘要

Obesity and overweight have become a major health concern, currently about 30% of the population suffers from obesity and this percentage is expected to increase to 50% by 2035. Obesity accounts for more than five million deaths since 2019 and its costs are about 2.19% of global gross domestic product (GDP). This paper focuses on how the utilization of Variational Autoencoders (VAEs) and the interpretation of the latent space can help in the predictions of the evolution of the patients using clinical data and 3D shape representation of the body. This research aims to bring treatment closer to patients and improve outcomes in the treatment of obesity and overweight without the need of expensive methods. The results show that the proposed method accurately reconstructs 3D models of the human body and clinical data, and the analysis of the latent space shows a new way to study a patient’s progress.