Uncovering and mitigating bias in large, automated MRI analyses of brain development
Safia Elyounssi1,2, Keiko Kunitoki1,2, Jacqueline A Clauss1,2
1Department of Psychiatry, Massachusetts General Hospital and Harvard Medical School.
Abstract:
Large, population-based MRI studies of adolescents promise transformational insights into neurodevelopment and mental illness risk 1,2. However, MRI studies of youth are especially susceptible to motion and other artifacts 3,4. These artifacts may go undetected by automated quality control (QC) methods that are preferred in high-throughput imaging studies, 5 and can potentially introduce non-random noise into clinical association analyses. Here we demonstrate bias in structural MRI analyses of children due to inclusion of lower quality images, as identified through rigorous visual quality control of 11,263 T1 MRI scans obtained at age 9-10 through the Adolescent Brain Cognitive Development (ABCD) Study6. Compared to the best-rated images (44.9% of the sample), lower-quality images generally associated with decreased cortical thickness and increased cortical surface area measures (Cohen's d 0.14-2.84). Variable image quality led to counterintuitive patterns in analyses that associated structural MRI and clinical measures, as inclusion of lower-quality scans altered apparent effect sizes in ways that increased risk for both false positives and negatives. Quality-related biases were partially mitigated by controlling for surface hole number, an automated index of topological complexity that differentiated lower-quality scans with good specificity at Baseline (0.81-0.93) and in 1,000 Year 2 scans (0.88-1.00). However, even among the highest-rated images, subtle topological errors occurred during image preprocessing, and their correction through manual edits significantly and reproducibly changed thickness measurements across much of the cortex (d 0.15-0.92). These findings demonstrate that inadequate QC of youth structural MRI scans can undermine advantages of large sample size to detect meaningful associations.
Insights
Rigorous quality control is crucial for youth neuroimaging studies. Poor quality MRI scans can introduce biases in brain structure analyses, affecting mental health research findings.
Area of Science:
- Neuroscience
- Medical Imaging
- Developmental Psychology
Background:
- Large-scale magnetic resonance imaging (MRI) studies in adolescents offer critical insights into brain development and mental health.
- Youth neuroimaging is particularly vulnerable to motion artifacts, which automated quality control (QC) may miss.
- Undetected artifacts can introduce bias into clinical association analyses.
Purpose of the Study:
- To investigate the impact of image quality on structural MRI analyses in children.
- To identify biases introduced by lower-quality scans in large youth cohorts.
- To evaluate methods for mitigating quality-related biases in neuroimaging data.
Main Methods:
- Visual quality control of 11,263 T1 MRI scans from the Adolescent Brain Cognitive Development (ABCD) Study (ages 9-10).
- Comparison of structural MRI measures (cortical thickness, surface area) between high-quality and lower-quality scans.
- Assessment of automated QC metrics, including surface hole number, for bias detection.
- Evaluation of manual edits on image preprocessing to correct topological errors.
Main Results:
- Lower-quality scans were associated with decreased cortical thickness and increased cortical surface area (Cohen's d 0.14-2.84).
- Inclusion of lower-quality scans distorted effect sizes, increasing the risk of false positives and negatives in association analyses.
- Surface hole number partially mitigated biases but did not eliminate them.
- Manual correction of subtle topological errors significantly altered cortical thickness measurements (d 0.15-0.92).
Conclusions:
- Inadequate quality control in youth structural MRI studies can compromise findings, even in large datasets.
- Rigorous visual QC and improved automated methods are essential for reliable neurodevelopmental research.
- Addressing image quality is critical for accurate associations between brain structure and clinical measures in adolescents.
