Login / Signup

Synthesis and validation algorithm followed by a weighting procedure to create a detailed anthropometric dataset for the German working-age population.

Alexander AckermannDominik BoninThomas JaitnerMarkus PetersDörte RadkeSascha Wischniewski
Published in: Ergonomics (2023)
For the German working-age population no publicly available and detailed anthropometric raw dataset exists, although several studies have collected anthropometric datasets. Unfortunately, the publication of raw data may be restricted by data usage regulations. This study presents a synthesis and validation algorithm to create a virtual copy of an already existing dataset. A detailed anthropometric dataset from a regional epidemiological public-health study in Germany was used for the synthesis and validation algorithm. Results revealed only minor deviations within the validation process. Compared to the original dataset, the virtual dataset was statistically almost identical. In a next step, the virtual dataset was weighted to approximate nationally representative values. In summary, the computed unweighted and weighted virtual data can be published without restrictions and used for ergonomic designing. Furthermore, the synthesis and validation algorithm is suitable for the generation of virtual copies and can be applied to other detailed anthropometric datasets.
Keyphrases
  • body composition
  • machine learning
  • public health
  • deep learning
  • electronic health record
  • big data
  • neural network
  • magnetic resonance imaging
  • randomized controlled trial
  • rna seq
  • network analysis