Related Experiment Video
Updated: Aug 7, 2026
Segmentation and Linear Measurement for Body Composition Analysis using Slice-O-Matic and Horos
Published on: March 21, 2021
Automated analysis of abdominal body composition using MRI: algorithm development and validation via CT comparison
1Department of Radiology, Seoul National University Hospital, Seoul National University College of Medicine, 101 Daehak-ro, Jongno-gu, Seoul 03080, Republic of Korea; Department of Radiology, Seoul National University College of Medicine, Seoul, Republic of Korea.
A new deep learning algorithm accurately segments abdominal body composition using MRI, showing strong correlation with CT scans. However, distinct reference values are needed due to systematic differences between MRI and CT measurements.
Area of Science:
- Radiology
- Medical Imaging
- Artificial Intelligence in Medicine
Background:
- Accurate body composition analysis is crucial for assessing metabolic health.
- Magnetic Resonance Imaging (MRI) offers a radiation-free alternative to Computed Tomography (CT) for body composition analysis.
- Automated segmentation of abdominal fat and muscle in MRI is challenging.
Purpose of the Study:
- To develop and validate a deep learning algorithm for automated segmentation of abdominal visceral fat (AVF), abdominal subcutaneous fat (ASF), and skeletal muscle (SM) in MRI data.
- To compare the effectiveness of MRI-based body composition analysis using the developed algorithm against traditional CT-based analysis.
Main Methods:
- A 3D nnU-Net deep learning model was developed for segmenting AVF, ASF, and SM in MRI scans.
- The model was trained on 105 whole-body MRI scans and validated on 67 abdominal MRI scans using Dice Similarity Coefficient (DSC).
- Cross-modality correlation and agreement were assessed between MRI (using the algorithm) and CT (using a commercial tool) in 54 patients via Pearson's correlation, ICC, and Bland-Altman analysis.
Main Results:
- The algorithm achieved high segmentation performance with mean DSCs ranging from 0.823 to 0.923 for different fat and muscle components.
- MRI-based measurements showed strong correlation (r>0.95) and excellent overall agreement (ICC>0.95) with CT measurements in both 3D and 2D analyses.
- However, absolute agreement between MRI and CT was less favorable, particularly for AVF, with wide limits of agreement.
Conclusions:
- The developed deep learning algorithm enables accurate MRI-based 3D segmentation of abdominal body composition.
- While showing strong correlation, systematic differences exist between MRI and CT measurements.
- Modality-specific reference values are necessary for the clinical application of MRI-based body composition analysis.
Related Concept Videos
01:10Computed Tomography
The technique was invented in the 1970s and is based on the principle that as X-rays pass through the body, they are absorbed or reflected at different levels. In the technique, a patient lies on a motorized platform while a computerized axial tomography (CAT) scanner rotates...
01:14Imaging Studies I: CT and MRI
Description of the Procedures
Computed Tomography (CT) scan:
Computed Tomography (CT) scans use X-ray technology to generate detailed images of bones, organs, and tissues. During the scan, the patient lies on a moving table...