Restoring statistical validity in group analyses of motion-corrupted MRI data
Antoine Lutti1, Nadège Corbin2,3, John Ashburner3
1Laboratory for Research in Neuroimaging, Department of Clinical Neurosciences, Lausanne University Hospital and University of Lausanne, Lausanne, Switzerland.
Abstract:
Motion during the acquisition of magnetic resonance imaging (MRI) data degrades image quality, hindering our capacity to characterise disease in patient populations. Quality control procedures allow the exclusion of the most affected images from analysis. However, the criterion for exclusion is difficult to determine objectively and exclusion can lead to a suboptimal compromise between image quality and sample size. We provide an alternative, data-driven solution that assigns weights to each image, computed from an index of image quality using restricted maximum likelihood. We illustrate this method through the analysis of quantitative MRI data. The proposed method restores the validity of statistical tests, and performs near optimally in all brain regions, despite local effects of head motion. This method is amenable to the analysis of a broad type of MRI data and can accommodate any measure of image quality.
Insights
Motion during magnetic resonance imaging (MRI) acquisition degrades image quality. This study introduces a data-driven weighting method to improve analysis validity and optimize sample size, outperforming traditional exclusion criteria.
Area of Science:
- Medical Imaging
- Neuroimaging
- Biostatistics
Background:
- Motion during MRI acquisition significantly degrades image quality.
- Current quality control methods involve image exclusion, which can be subjective and reduce sample size.
- This compromises the ability to accurately characterize diseases in patient populations.
Purpose of the Study:
- To develop and validate a data-driven method for handling motion-affected MRI data.
- To provide an objective alternative to image exclusion in MRI analysis.
- To improve statistical validity and optimize the balance between image quality and sample size.
Main Methods:
- A novel method assigning weights to each MRI image based on an image quality index.
- Utilized restricted maximum likelihood (REML) for weight computation.
- Applied the method to quantitative MRI data analysis.
Main Results:
- The proposed weighting method restores the validity of statistical tests.
- The approach performs near-optimally across all brain regions, even with localized motion artifacts.
- Demonstrated effectiveness in quantitative MRI data analysis.
Conclusions:
- This data-driven weighting approach offers a superior alternative to image exclusion for motion-affected MRI data.
- The method is broadly applicable to various MRI data types and quality measures.
- Enhances the reliability and efficiency of neuroimaging research in clinical populations.
