Related Experiment Video
Updated: Jun 16, 2025
05:55
Modeling the Functional Network for Spatial Navigation in the Human Brain
Published on: October 13, 2023
1.0K
Fast Whole-Brain MR Multi-Parametric Mapping with Scan-Specific Self-Supervised Networks
Amir Heydari1, Abbas Ahmadi1, Tae Hyung Kim2
1Department of Industrial Engineering and Management Systems, Amirkabir University of Technology, Tehran, Iran.
Arxiv
|August 16, 2024
Summary
Accelerated quantitative MRI mapping is now faster and more accurate. A new Joint MAPLE technique significantly reduces scan times for tissue parameter quantification, improving diagnostic capabilities.
Area of Science:
- Medical Imaging
- Biophysics
- Machine Learning in Medicine
Background:
- Quantitative MRI (qMRI) offers powerful diagnostic insights but is limited by long scan times.
- Accelerated techniques using parallel imaging, modeling, and deep learning show promise but face limitations in speed and map quality.
- Joint MAPLE is a state-of-the-art technique for multi-parametric mapping but has lengthy reconstruction times.
Purpose of the Study:
- To develop a significantly faster version of the Joint MAPLE technique for quantitative MRI.
- To maintain or improve the mapping performance of Joint MAPLE while drastically reducing reconstruction time.
- To enable practical, high-resolution, scan-specific quantitative MRI parameter mapping.
Main Methods:
- Developed a faster Joint MAPLE framework by synergistically combining coil compression, random slice selection, parameter-specific learning rates, and transfer learning.
- Applied the framework to multi-echo, multi-flip angle (MEMFA) datasets for joint mapping of T1, proton density, and field inhomogeneity.
- Evaluated reconstruction time reduction and mapping accuracy compared to original Joint MAPLE and other state-of-the-art methods.
Main Results:
- Achieved up to a 700-fold speed-up in reconstruction time compared to the original Joint MAPLE.
- Reduced whole-brain MEMFA dataset processing time from ~260 hours to an average of 21 minutes.
- Demonstrated approximately 2-fold improvement in mapping performance (lower root mean squared error) over standard and state-of-the-art techniques.
Conclusions:
- The proposed framework dramatically accelerates Joint MAPLE reconstruction for quantitative MRI parameter mapping.
- This advancement makes high-quality, scan-specific quantitative MRI more feasible for routine clinical and research applications.
- The technique offers superior mapping accuracy and efficiency, overcoming previous adoption barriers for quantitative MRI.
