Skip to navigation Skip to main content Skip to footer

Approved Research

Developing and Validating a Web-Based Tool for Automatic Body Composition Analysis with abdominal MRI and DXA scans

Principal Investigator: Professor Mirza Faisal Beg
Approved Research ID: 43544
Approval date: January 25th 2021

Lay summary

Developing a web-based tool to automatically analyze the body composition on MRI/DXA images. A model is trained with the sample MRI/DXA images that have been manually segmented. This research project includes different steps that are independently useful for future researches as well.

Developing a tissue segmentation protocol for MRI/DXA provides the labelled data for training the model leveraging the machine learning methods as well as a source for further research and segmentation protocols.

Designing an automatic method to generate the segmentation for the fat and muscle regions on MRI/DXA images. These segmentations will further be used for the estimation of body composition. There are many applications including, but not limited to cancer drug dose estimation, for the information extracted from body composition analysis. Other fields studying the patients suffer from degenerative loss of skeletal muscle, muscle function/performance and heart diseases can also benefit from the automatic body composition analysis.

The aim of this research is to provide the health clinics and physicians with a smart and user-friendly tool to be able to make the crucial clinical decisions with the quick and accurate results produced with this web-based tool.