Inter- and intra-scanner variations in four magnetic resonance imaging image quality parameters

Area of Science:

• Medical Imaging
• Radiology
• Quality Assurance

Background:

• Quality assurance (QA) protocols for magnetic resonance imaging (MRI) scanners often include phantom checks to ensure equipment constancy.
• Daily QA images offer potential for studying scanner baseline performance and image quality variations.
• Automatic error detection can be implemented with established baselines and variation profiles.

Purpose of the Study:

• To investigate inter-scanner variability in MRI image quality parameters.
• To establish baseline performance and variation profiles for MRI scanners.
• To evaluate the efficacy of automatic detection of abnormal QA images compared to human observation.

Main Methods:

• Monitored four image quality parameters (SNR, uniformity, ghosting, geometrical distortions) across 17 MRI scanners for six months.
• Determined baseline levels and normal variation ranges for each parameter.
• Compared automatic abnormal image detection with visual assessment by human observers.

Main Results:

• Significant inter-scanner differences were observed in all QA parameters.
• Some QA results exceeded established tolerances, indicating potential equipment issues.
• No clear relationship was found between scanner field strength or mobility and QA results.

Conclusions:

• MRI scanner baseline performance and image quality parameter variations differ substantially between units.
• Scanner-specific error thresholds, derived from parameter means and standard deviations, are effective for detecting abnormal QA images.
• Automatic analysis of daily QA images can enhance MRI quality control.