Automated Quality Evaluation Index for Arterial Spin Labeling Derived Cerebral Blood Flow Maps

Area of Science:

• Neuroimaging
• Medical Physics
• Radiology

Background:

• Arterial spin labeling (ASL) derived cerebral blood flow (CBF) maps can suffer from artifacts and noise, impacting image quality.
• Objective assessment of ASL CBF map quality is crucial for reliable clinical interpretation.

Purpose of the Study:

• To develop and validate an automated, objective quality evaluation index (QEI) for ASL CBF maps.
• To establish a scalable method for quality control in ASL neuroimaging.

Main Methods:

• Developed a QEI using 101 2D ASL CBF maps rated by neuroradiologists.
• Validated the QEI through cross-validation, correlation with CBF reproducibility, effect size analysis, and comparison with manual ratings and an existing automated metric.
• Utilized data from 221 adults, including patients with Alzheimer's disease, Parkinson's disease, and traumatic brain injury.

Main Results:

• The automated QEI demonstrated strong correlation with manual quality ratings (R=0.83 for 2D, R=0.86 for 3D ASL).
• QEI showed significant inverse correlation with CBF reproducibility (R=-0.74) and improved effect sizes when low-quality data was excluded.
• The developed QEI outperformed a previously existing automated quality metric.

Conclusions:

• The automated QEI provides a reliable and scalable method for quality control of ASL CBF maps.
• This automated approach achieves performance comparable to expert manual ratings.
• The QEI can enhance the consistency and accuracy of ASL-based neuroimaging analyses.