Towards individualized cortical thickness assessment for clinical routine

Area of Science:

• Neuroimaging
• Biomarkers
• Neurology

Background:

• Cortical thickness, derived from T1-weighted MRI, is a key neuroimaging biomarker.
• While correlated with neurological and psychiatric conditions, its clinical use is limited by a lack of standardized reference systems for individual data.
• No established method exists to rate individual cortical thickness against control populations for routine clinical application.

Purpose of the Study:

• To develop and validate methods for assessing the statistical significance of cortical thinning (atrophy) in individual subjects.
• To establish a reference system for rating individual cortical thickness data against control populations.
• To improve the clinical utility of cortical thickness measurements for diagnosis and prognosis.

Main Methods:

• Compared nonparametric methods for assessing statistical significance of cortical thinning.
• Utilized Human Connectome Project (HCP) data (n=1000) for null distribution calculation and (n=113) for sensitivity/specificity analysis.
• Simulated atrophy to assess method sensitivity and calculated validation measures cumulatively and regionally.

Main Results:

• The optimal method involved generating null distributions for distinct brain regions based on extreme population values.
• Achieved high cumulative specificity (98.9%) and sensitivity (80%) for simulated atrophy of 23%.
• Demonstrated regional variations in atrophy detection, supporting differential diagnosis.

Conclusions:

• Validated rating of individual cortical thickness measures is now feasible for clinical routine.
• Enables early detection of atrophy signs on MRI, potentially before visual identification.
• Regional validation facilitates the detection of distinct atrophy patterns, enhancing differential diagnosis and prognosis.