Censoring distances based on labeled cortical distance maps in cortical morphometry

Area of Science:

• Neuroimaging
• Computational Anatomy
• Psychiatry

Background:

• Neuropsychiatric disorders are associated with structural brain differences.
• Labeled Cortical Distance Mapping (LCDM) quantifies cortical morphometry using distances from gray matter (GM) to white matter (WM) surfaces.
• Standard LCDM measures like volume may not capture all morphometric information.

Purpose of the Study:

• To develop and validate a method for extracting more detailed morphometric information from LCDM data.
• To investigate the influence of data aggregation in censored LCDM analysis.
• To identify specific locations of cortical morphometric differences in major depressive disorder (MDD).

Main Methods:

• Introduced censored LCDM, partitioning distance ranges and retaining distances below a threshold.
• Conducted extensive Monte Carlo simulations to assess the impact of data aggregation.
• Applied the method to ventral medial prefrontal cortex (VMPFC) gray matter in MDD, high-risk (HR), and control (Ctrl) subjects.

Main Results:

• Censored LCDM provides location-specific morphometric information beyond pooled distances.
• Monte Carlo simulations confirmed that data aggregation influence is negligible.
• Observed significant reductions in VMPFC laminar thickness in MDD and HR subjects compared to controls.
• Determined the specific GM LCDM distances where VMPFC thickness differences initiate.

Conclusions:

• Censored LCDM is a valuable extension of LCDM for detailed morphometric analysis in neuropsychiatric disorders.
• The method accurately identifies localized cortical thinning in MDD.
• This methodology is applicable to other cortical structures and diseases.