Automated MRI Analysis for Abdominal Adipose Tissue Measurement
Computing & Wireless : Application Software
Available for licensing
- Bugao Xu, Ph.D. , Division of Textiles and Apparel
- Jingjing Sun , Human Ecology
Traditionally, MRI imaging software requires human intervention and thus introduces observer bias. Researchers at The University of Texas at Austin have developed an unsupervised algorithm that substantially increases the reliability, objectivity, and efficiency of MRI abdominal fat assessment by avoiding manual operations and observer bias. This software has been shown to effectively correct inhomogeneity commonly noticeable in T1-weighted MRI images, and thus enhances the accuracy of adiposity assessment as well.
This software provides a fully-automated assessment of abdominal adipose tissues. It identifies fat tissue voxels and differentiates visceral fat and subcutaneous fat tissue types without human intervention and prior knowledge. It contains programs that perform abdominal mask extraction, noise reduction, intensity inhomogeneity correction, image clustering and segmentation. This software is suitable for MRI data acquired by more advanced imaging modalities, such as water-saturation imaging and IDEAL (Iterative Decomposition with Echo Asymmetry and Least-squares estimation) technique.
- Reliability: It significantly increases the reliability by eliminating observer bias.
- Cost: Similar software on market is expensive; for example, sliceOmatic costs $4,000 per license.
- Automation: It can be used by non-experts.
- Processing time: It takes only ~25 seconds per slice and it is suitable for assessing large populations.
- Prior knowledge: It does not need prior knowledge, and thus broadens its application to the data acquired by different imaging modalities.
MRI slice analysis for measuring visceral and subcutaneous adipose tissues.
The technology encompasses a wide range of applications, such as visualization and analysis of MR images, assessment of size, volume and shape of tissue compartments (e.g. fat and muscle), manual modification of processed fat assessment, image quality improvements (especially the inhomogeneous T1-weighted MRI), tissue differentiation and examination, 3D-reconstruction of MRI slices, abdominal obesity evaluation for the determination of medication dosage, body fat estimate in human body composition research, assessing and monitoring obesity for public health, monitoring inner adipose tissue growth or reduction of overweight patients, etc.