Compressed Sensitivity Encoding Artificial Intelligence Accelerates Brain Metastasis Imaging by Optimizing Image...

Area of Science:

• Medical Imaging
• Artificial Intelligence in Radiology
• Neuro-oncology

Background:

• Accelerating MRI acquisition speed without compromising image quality is a significant challenge.
• Contrast-enhanced (CE) 3D T1WI and CE 3D-FLAIR sequences are crucial for detecting brain metastases (BM).

Purpose of the Study:

• To evaluate the feasibility of CS-AI for reconstructing CE 3D T1WI and CE 3D-FLAIR sequences for BM detection.
• To determine the optimal acceleration factor (AF) for CS-AI in clinical BM imaging.

Main Methods:

• Fifty-one patients with suspected BM were included in the study.
• CE 3D-T1WI and CE 3D-FLAIR sequences were reconstructed using CS-AI with varying AFs.
• Compressed SENSE encoding acceleration 6 (CS6) served as the reference standard.
• Quantitative (SNR, CNR) and qualitative assessments were performed by neuroradiologists.

Main Results:

• CS-AI protocols demonstrated superior CNR and SNR compared to the standard CS protocol.
• CS-AI achieved good image quality up to AF 10, outperforming CS6.
• The CS-AI10 protocol provided optimal image quality, enhancing delineation of anatomical structures and lesions.
• CS-AI10 reduced scan times by approximately 40% for both sequences.

Conclusions:

• CS-AI offers a more effective reconstruction approach than conventional CS for BM detection using CE 3D-T1WI and CE 3D-FLAIR sequences.
• The CS-AI10 protocol is clinically suitable, balancing optimal image quality with reduced scan time.