Europeanhhm

Deep Learning to Estimate Impaired Glucose Metabolism From Magnetic Resonance Imaging of the Liver: an Opportunistic Population Screening Approach

Lea J. Michel, Susanne Rospleszcz, Marco Reisert, Alexander Rau, Johanna Nattenmueller, Wolfgang Rathmann, Christopher. L. Schlett, Annette Peters, Fabian Bamberg, Jakob Weiss

Abstract

Diabetes is a global health challenge, and many individuals are undiagnosed and not aware of their increased risk of morbidity/mortality although dedicated tests are available, which indicates the need for novel population-wide screening approaches. Here, we developed a deep learning pipeline for opportunistic screening of impaired glucose metabolism using routine magnetic resonance imaging (MRI) of the liver and tested its prognostic value in a general population setting.

Introduction

Type 2 diabetes is defined as a relative insulin deficiency resulting from defects in the pancreatic betta cells and insulin resistance in target organs [1]. During the past decades, the prevalence has dramatically increased due to rapid urbanisation and a more sedentary lifestyle [2,3,4]. Currently, the global prevalence of diabetes counts 536.6 million people or 10.5% of the adult population and is considered to further increase with an estimate of 783.2 million diseased individuals by 2045 [4]. This is of great importance, as the management of type 2 diabetes and its complications is placing a substantial economic burden on the healthcare system [5]. For example, Jacobs et al. revealed, that in Germany,

Materials and method

All analyses were performed using data from the Cooperative Health Research in the Region of Augsburg (KORA) MRI study, a cross-sectional MRI study nested in the prospective epidemiological cohort of the KORA main study [18,19]. The study population of the KORA-MRI study was recruited between June 2013 and September 2014 and consisted of 400 participants who underwent a multiparametric whole-body MRI study protocol as previously described in detail [19]. Subjects were eligible if they met the following inclusion criteria: consent to undergo whole-body MRI, no prior cardiovascular disease, and a standardized assessment of glucose metabolism.

Results

An overview of the study design is presented in Fig 1. Of the 400 participants included in the KORA-MRI study, 46 participants were excluded due to missing or corrupted imaging data. Additional 15 participants were excluded due to missing covariates resulting in a final study cohort of 339 individuals (Fig 2) with a mean age of 56.3±9.1 years. 58.1% were male and 23.3% (79/339) vs. 13.6% (46/339) were classified as having prediabetes and type 2 diabetes, respectively. Participants with type 2 diabetes were significantly older, more likely men and presented more often with prevalent hypertension and hepatic steatosis compared to normoglycemic controls (all p<0.001).

Discussion

In this study, we developed a fully automatic deep learning framework to investigate the association between radiomic shape features of the liver and impaired glucose metabolism. Our results demonstrate that MRI-derived radiomic shape features of the liver are independently associated with impaired glucose metabolism after adjustment for traditional cardiometabolic risk factors and hepatic steatosis in a cohort of individuals with prediabetes, type 2 diabetes and normal controls.

Acknowledgments

We thank all participants for their long-term commitment to the KORA study, the staff for data collection and research data management and the members of the KORA Study Group (https://www.helmholtz-munich.de/en/epi/cohort/kora) who are responsible for the design and conduct of the study.

Citation: Michel LJ, Rospleszcz S, Reisert M, Rau A, Nattenmueller J, Rathmann W, et al. (2024) Deep learning to estimate impaired glucose metabolism from Magnetic Resonance Imaging of the liver: An opportunistic population screening approach. PLOS Digit Health 3(1): e0000429. https://doi.org/10.1371/journal.pdig.0000429

Editor: Martin G. Frasch, University of Washington, UNITED STATES

Received: July 15, 2023; Accepted: December 7, 2023; Published: January 16, 2024

Copyright: © 2024 Michel et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.

Data Availability: The informed consent given by KORA study participants does not cover data posting in public databases. However, data are available upon request by means of a project agreement. Requests should be sent to kora.passt@helmholtz-muenchen.de and are subject to approval by the KORA Board.

Funding: The KORA study was initiated and financed by the Helmholtz Zentrum München – German Research Center for Environmental Health, which is funded by the German Federal Ministry of Education and Research (BMBF) and by the State of Bavaria. Data collection in the KORA study is done in cooperation with the University Hospital of Augsburg. This study was funded by German Research Foundation (Bonn, Germany) grant BA 4233/4-1. For this sub-study the authors received no specific funding. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.

Competing interests: The authors have declared that no competing interests exist.

Source: https://journals.plos.org/digitalhealth/article?id=10.1371/journal.pdig.0000429#ack

 

Equinix accelerated medical research...The Healthcare Patient Experience & Engagement Summit 2024Healthcare Innovation & Transformation SummitHealthcare CNO SummitHealthcare CMO Summit