| Title: | Joint analysis of abdominal and cerebral morphology in diabetes |
| Lead Institution: | Ludwig Maximilian University of Munich |
| Principal investigator: | Professor Christian Wachinger |
Application 34479
WARNING: the interactive features of this website use CSS3, which your browser does not support. To use the full features of this website, please update your browser.
About
The aim is to jointly analyze abdominal and cerebral morphology in diabetes. Most studies to date have separately analyzed associations between diabetes and the morphology of the abdomen and the brain, while we want to take advantage of the UK biobank whole-body imaging protocol and study them jointly. We will further integrate genetic risk loci for diabetes in the analysis, to investigate their impact on morphology. For quantifying morphology, we will be working with spectral shape descriptors. These present a high-dimensional characterization of a structure and can therefore retain more of the geometric information than volume-based representations. The purpose of the project is to gain a better understanding of morphological changes in the abdomen and brain in diabetes. A substantial part of the project will be about developing new methods for accurately characterizing body morphology. Such methods can also be applied to studying other diseases. The contributions in this project are therefore not only applicable to diabetes but also to studying general questions in the population. Further, studying morphological changes in diabetes with the associated genetic risk can help in developing biomarkers for the early prediction, which is of wide public interest due to the high prevalence. For quantifying body morphology, we will access MRI brain scans and whole-body scans. In the first step, we will automatically identify brain structures and organs in these scans by extending our work on segmentation with deep learning. In the second step, we will use shape descriptors to transform the segmentation masks into vectors that represent the body morphology. In the third step, we will jointly analyze the relation between diabetes and abdominal and cerebral morphology, while integrating the genetic risk in the model. Full cohort3 Publications
| Pub ID | Title | Author(s) | Year | Journal |
|---|---|---|---|---|
| 9331 | Abdominal organ segmentation via deep diffeomorphic mesh deformations | Fabian Bongratz (+2) | 2023 | Scientific Reports |
| 8359 | Deep learning for the prediction of type 2 diabetes mellitus from neck-to-knee Dixon MRI in the UK biobank | Christian Wachinger (+2) | 2023 | Heliyon |
| 9268 | Detect and correct bias in multi-site neuroimaging datasets | Christian Wachinger (+3) | 2020 | Medical Image Analysis |
Enabling scientific discoveries that improve human health