Automatic quality assessment in structural brain magnetic resonance imaging

Area of Science:

• Medical Imaging
• Radiology
• Biomedical Engineering

Background:

• Magnetic Resonance Imaging (MRI) is crucial for medical diagnosis.
• Image artifacts degrade MRI reliability, challenging human and automated diagnoses.
• Assessing 3D structural MRI quality is essential for accurate interpretation.

Purpose of the Study:

• To develop a fully-automatic method for measuring 3D structural MRI image quality.
• To identify and quantify image degradation from sources like motion, blurring, and ghosting.
• To provide a tool for enhancing diagnostic reliability and clinical workflow.

Main Methods:

• Analyzing the air background of magnitude images in 3D structural MRI.
• Deriving quality measures sensitive to common MRI artifacts.
• Validating the method on a large dataset of 749 T1-weighted head scans from the ADNI study.

Main Results:

• The automated quality assessment method demonstrated high sensitivity and specificity (>85%) compared to expert ratings.
• Quality indices were independent of MRI system hardware and software variations.
• The method effectively detected image degradation from bulk motion, incomplete spoiling, blurring, and ghosting.

Conclusions:

• The proposed automated quality assessment is valuable for research and clinical MRI.
• It can improve workflow by identifying scans needing reacquisition before patient departure.
• This method enhances the reliability of MRI-based diagnoses and computer-aided detection systems.