Automated Analysis for MR Coil QA

Area of Science:

• Medical Imaging Physics
• Radiology Quality Control
• Software Development in Healthcare

Background:

• Phantom-based quality assurance (QA) is crucial for magnetic resonance imaging (MRI) coils, assessing signal-to-noise ratio (SNR), uniformity, and homogeneity.
• Manual slice selection and region-of-interest (ROI) placement in current QA workflows introduce variability and inefficiencies.

Purpose of the Study:

• To develop a unified, fully automated desktop-based workflow for MRI coil QA.
• To eliminate manual slice selection and ROI placement, reducing operator dependency.

Main Methods:

• A Python desktop application was developed to automate MRI coil QA for various coil types (body, torso, head-and-neck).
• The workflow automatically performs slice selection, phantom detection, and ROI placement using DICOM and free-induction decay data.
• Standard QA metrics including SNR, percent image uniformity, and magnetic field homogeneity are calculated automatically.

Main Results:

• The automated workflow successfully executed all QA tasks without manual intervention.
• Consistent QA metrics were generated for different coil types and analyses (weekly QA, magnetic field homogeneity).
• Standardized outputs enabled objective coil and channel performance evaluation, minimizing subjective bias.

Conclusions:

• An automated MRI QA workflow was developed, utilizing standardized testing methods.
• The workflow delivers consistent results and reduces human error in clinical quality control.
• This automation enhances the reliability and efficiency of MRI coil QA procedures.