DLA based compressed sensing for high resolution MR microscopy of neuronal tissue

Area of Science:

• Magnetic Resonance Imaging
• Biomedical Engineering
• Neuroscience

Background:

• Compressed sensing (CS) enables faster MRI by reconstructing images from undersampled data.
• Traditional CS undersampling methods often rely on polynomial probability density functions.
• High-field preclinical MRI scanners offer advanced imaging capabilities but require efficient acquisition strategies.

Purpose of the Study:

• To implement and evaluate a novel compressed sensing (CS) undersampling trajectory based on the diffusion limited aggregation (DLA) random growth model on a 17.2 T preclinical scanner.
• To compare the performance of the DLA-based CS undersampling with traditional polynomial probability density function-based undersampling.
• To assess the applicability of this DLA-CS method for imaging live neuronal tissues and its impact on acquisition time and image quality for automated cell segmentation.

Main Methods:

• Implementation of a diffusion limited aggregation (DLA) random growth model for generating undersampling trajectories.
• Application of the DLA-based CS method on a high-field (17.2 T) preclinical MRI scanner.
• Testing the method on a library of images and live neuronal tissues.
• Utilizing an automatic cell segmentation algorithm to evaluate image quality.

Main Results:

• The DLA-based CS undersampling trajectory demonstrated superior performance compared to traditional polynomial probability density function-based undersampling on a tested image library.
• The DLA-CS method was successfully applied to imaging live neuronal tissues.
• Significantly shorter acquisition times were achieved while preserving image quality sufficient for automatic neuron identification.

Conclusions:

• The diffusion limited aggregation (DLA) trajectory offers an effective strategy for compressed sensing (CS) in high-field preclinical MRI.
• This novel undersampling approach improves imaging efficiency for neuronal tissues, enabling faster scans without compromising essential image quality for cellular analysis.