Being overweight and having a lot of body fat increase the risk of diabetes. However, not every overweight person also develops the disease. The decisive factor is where the fat is stored in the body. If fat is stored under the skin, it is less harmful than fat in deeper areas of the abdomen (known as visceral fat). How fat is distributed throughout the body can be easily visualized with whole-body magnetic resonance imaging. “We have now investigated whether type 2 diabetes could also be diagnosed on the basis of certain patterns of body fat distribution using MRI,” said last author Prof. Robert Wagner, explaining the researchers’ approach.
Deep learning trained with over 2000 MRI scans
To detect such patterns, the researchers used artificial intelligence (AI). They trained deep learning (machine learning) networks with whole-body MRI scans of 2,000 people who had also undergone screening with the oral glucose tolerance test (abbreviated OGTT). The OGTT can screen for impaired glucose metabolism and diagnose diabetes. This is how the AI learned to detect diabetes.
Lower abdominal fat accumulation an important indicator of diabetes pathogenesis
“An analysis of the model results showed that fat accumulation in the lower abdomen plays a crucial role in diabetes detection,” Wagner said. Further additional analysis also showed that a proportion of people with prediabetes, as well as people with a diabetes subtype that can lead to kidney disease, can also be identified via MRI scans.
The researchers are now working to decipher the biological regulation of body fat distribution. One goal is to identify the causes of diabetes through new methods such as the use of AI in order to find better preventive and therapeutic options.
Dietz et al.: Diabetes detection from whole-body magnetic resonance imaging using deep learning. JCI Insight, DOI: https://doi.org/10.1172/jci.insight.146999
*Deep learning is a special method from the field of machine learning with artificial neural networks (ANN) and thus also a sub-area of artificial intelligence (AI). Deep learning is particularly suitable when there is a lot of unstructured data – such as images and scans. In order to teach deep learning algorithms to correctly evaluate images and predict diagnoses, they are trained on annotated (labeled) data.
The German Center for Diabetes Research (DZD) e.V. is one of six German Centers for Health Research. It brings together experts in the field of diabetes research and combines basic research, epidemiology and clinical application. By adopting an innovative, integrative approach to research, the DZD aims to make a substantial contribution to the successful, personalized prevention, diagnosis and treatment of diabetes mellitus. The members of the association are Helmholtz Zentrum München – German Research Center for Environmental Health, the German Diabetes Center (DDZ) in Düsseldorf, the German Institute of Human Nutrition (DIfE) in Potsdam-Rehbrücke, the Institute of Diabetes Research and Metabolic Diseases of Helmholtz Zentrum München at the University of Tübingen, and the Paul Langerhans Institute Dresden of Helmholtz Zentrum München at the University Medical Center Carl Gustav Carus of TU Dresden, associated partners at the universities in Heidelberg, Cologne, Leipzig, Lübeck and Munich, and other project partners. Further information: www.dzd-ev.de/en
Founded in 1805, Tübingen University Hospital is one of the leading centers of German university medicine. As one of the 33 university hospitals in Germany, it contributes to the successful combination of high-performance medicine, research and teaching. Well over 400,000 inpatients and outpatients from all over the world benefit annually from this combination of science and practice. The clinics, institutes and centers unite all specialists under one roof. The experts work together across disciplines and offer each patient the best possible treatment based on the latest research findings. Tübingen University Hospital conducts research for better diagnoses, therapies and healing chances; many new treatment methods are clinically tested and applied here. In addition to diabetology, neuroscience, oncology, immunology, infection research and vascular medicine are research priorities in Tübingen. The Department of Diabetology /Endocrinology has been the center of interdisciplinary research over the past 25 years, especially with the participation of surgery, radiology and laboratory medicine. This important discovery of the prediabetes subtypes was only possible due to the interdisciplinary collaboration between the hospital’s various departments. Tübingen University Hospital is a reliable partner in four of the six German Centers for Health Research initiated by the German Federal Government. www.medizin.uni-tuebingen.de
Method of Research
Subject of Research
Diabetes detection from whole-body magnetic resonance imaging using deep learning
Article Publication Date
Disclaimer: AAAS and EurekAlert! are not responsible for the accuracy of news releases posted to EurekAlert! by contributing institutions or for the use of any information through the EurekAlert system.