Berlin 2024 – scientific programme

Parts | Days | Selection | Search | Updates | Downloads | Help

ST: Fachverband Strahlen- und Medizinphysik

ST 5: Poster Session

ST 5.3: Poster

Wednesday, March 20, 2024, 11:00–12:30, Poster A

Unravel patterns of chronic diseases in India using Graph Networks — •Carola Behr, Anna Nitschke, Jannis Demel, Jonathan Berthold, Carlos Brandl, and Matthias Weidemüller — Physikalisches Institut, Universität Heidelberg, Deutschland

To uncover insights from complex and diverse medical datasets, Network Science and Graph based approaches can play an important role. Based on the DHS 2019-2021 dataset from India, we apply different methods to gain a deeper understanding of the prevalence of chronic diseases. In this poster we present the application of techniques such as Community Detection, which aims to identify inherent structures within the dataset, and Graph Convolutional Networks, designed to exploit the rich relational information embedded in the data. Our goal is first to uncover the nuanced connections of socio-economical parameters and biological parameters within that dataset using advanced network-centric methodologies and second improve the disease management for individuals and the health care system.

Keywords: Graph Neural Networks; Network Science; Chronic Diseases; Unsupervised