Intelligent Neurovascular Imaging Engine (INIE): Topology-Aware Compressed Sensing and Multimodal Super-Resolution...

Area of Science:

• Medical Imaging
• Neuroscience
• Biomedical Engineering

Background:

• Conventional MRI workflows for acute cerebrovascular disorders are limited by speed, motion sensitivity, and lack of physiological context.
• Rapid and reliable neuroimaging is crucial for timely diagnosis and treatment in time-sensitive neurological conditions.

Purpose of the Study:

• To introduce the Intelligent Neurovascular Imaging Engine (INIE), a novel framework for accelerated and physiologically consistent neurovascular MRI.
• To optimize accelerated data acquisition, physics-grounded reconstruction, and cross-scale physiological consistency using sensor information and topology awareness.

Main Methods:

• INIE integrates adaptive sampling, structured low-rank priors, and topology-preserving objectives with multimodal physiological sensors and scanner telemetry.
• The framework employs phase-consistent gating and confidence-weighted reconstruction under realistic operating conditions.
• Evaluated using synthetic phantoms, a porcine stroke model, and retrospective human datasets.

Main Results:

• INIE achieved over 70% acquisition acceleration with high reconstruction fidelity (PSNR ≈35-36 dB, SSIM ≈0.90-0.92).
• Topology-aware analysis demonstrated a twofold reduction in Betti number deviation compared to baseline methods.
• Improved large-vessel occlusion detection accuracy to ~93% and reduced time-to-decision to under three minutes.

Conclusions:

• Sensor-informed, topology-aware, closed-loop imaging enhances the reliability and physiological consistency of accelerated neurovascular MRI.
• INIE supports faster, more robust decision-making in acute cerebrovascular imaging workflows.
• This approach holds significant promise for improving patient outcomes in acute stroke care.