MRI-quality and morphometric MRI analysis to identify focal cortical dysplasia: An exploratory study

Area of Science:

• Neuroimaging
• Epilepsy Research
• Medical Image Analysis

Background:

• Morphometric Analysis Program v2018 (MAP18) assists in detecting subtle focal cortical dysplasia (FCD) in normal MRI scans for epilepsy surgery evaluation.
• Investigated the impact of reduced MRI signal-to-noise ratio (SNR) and spatial resolution (SR) on MAP18's FCD detection capabilities.

Purpose of the Study:

• To assess how diminished MRI quality (SNR and SR) affects MAP18's ability to detect focal cortical dysplasia (FCD).
• To identify potential limitations of MAP18 in identifying FCD under suboptimal imaging conditions.
• To inform strategies for improving surgical planning and patient outcomes in focal epilepsy.

Main Methods:

• Retrospective analysis of 30 MRI scans with radiologically confirmed FCD.
• Artificial reduction of SNR and SR in ten incremental steps.
• Multilevel analysis to evaluate the impact of reduced SNR and SR on MAP18 detection metrics (z-score, volume, cluster count).

Main Results:

• Reduced SR and SNR significantly affected FCD cluster z-score, volume, and the number and volume of detected clusters.
• FCD became undetectable by MAP18 after six SR reduction steps (2.8 × 2.8 × 2.8 mm³ voxel size) and two SNR reduction steps.
• SNR reduction also significantly impacted the z-score of the total detected cluster count.

Conclusions:

• Reduced spatial resolution and signal-to-noise ratio negatively impact FCD detection using MAP18 post-processing.
• MAP18 evaluators must assess MRI quality, noting that voxel sizes of 2.8 × 2.8 × 2.8 mm³ and higher, or significant visual noise, may impede FCD detection.
• Ensuring high-quality MRI scans is crucial for accurate FCD detection and successful epilepsy surgery outcomes, as repeating low-quality scans is preferable to missed diagnoses.