In Vivo Imaging With a Low-Cost MRI Scanner and Cloud Data Processing in Low-Resource Settings

Area of Science:

• Medical Imaging
• Biomedical Engineering
• Low-Field MRI

Background:

• Low-resource settings face significant challenges in accessing advanced medical imaging technologies like Magnetic Resonance Imaging (MRI).
• Existing low-field MRI systems often suffer from operational limitations due to hardware constraints and environmental factors in these regions.

Purpose of the Study:

• To demonstrate in vivo imaging capabilities using a low-cost, low-field MRI scanner developed and operated in Africa.
• To showcase how systematic hardware and software enhancements can overcome operational limitations in low-resource environments.

Main Methods:

• Upgraded a 46-mT Halbach MRI scanner with improved grounding, shielding, control electronics, and open-source software.
• Quantified noise performance and acquired 3D brain images using RARE sequences.
• Implemented cloud-based reconstructions for distortion correction using magnetic field maps.

Main Results:

• Achieved noise levels below three times the thermal limit, indicating improved signal-to-noise ratio.
• Demonstrated stable scanner operation over multi-day measurements.
• Successfully acquired and distortion-corrected 3D T1- and T2-weighted brain images with remote GPU processing.

Conclusions:

• Low-cost MRI systems can attain clinically relevant image quality by mitigating electromagnetic noise and power grid instabilities.
• Highlights the feasibility of sustainable MRI development and deployment in low-resource settings.
• Identifies stable power delivery and local capacity building as crucial for clinical translation.