PUNCH: Population Characterization of Heterogeneity

Area of Science:

• Neuroscience
• Clinical Psychology
• Biostatistics

Background:

• Neuropsychiatric disorders exhibit significant heterogeneity, complicating biomarker discovery and objective population description.
• Existing methods struggle to capture complex sample characteristics using a single, continuous measure.
• There is a critical need for reliable tools to quantify clinical heterogeneity.

Purpose of the Study:

• To introduce and validate a novel method, Population Characterization of Heterogeneity (PUNCH), for quantifying clinical population heterogeneity.
• To demonstrate PUNCH's ability to integrate diverse phenotypic scores into a single, interpretable severity measure.
• To apply PUNCH to Autism Spectrum Disorder (ASD) and examine its utility in neuroimaging research.

Main Methods:

• Developed PUNCH as a decision-level fusion technique incorporating multiple phenotypic scores with interpretable weights.
• Applied the PUNCH framework to simulated datasets and a large cohort of youth with ASD.
• Stratified PUNCH scores within the ASD sample to analyze group differences in diffusion-weighted brain imaging data.

Main Results:

• PUNCH successfully quantified heterogeneity in simulated and real-world clinical datasets.
• In youth with ASD, PUNCH scores revealed that more severely affected subgroups showed expanded differences in brain imaging compared to controls.
• Severity, as measured by PUNCH, moderated group differences in diffusion-weighted imaging findings.

Conclusions:

• PUNCH provides a reliable measure for quantifying clinical sample heterogeneity.
• The PUNCH tool can be utilized to generate robust severity assessments that account for population variability.
• Findings highlight the utility of PUNCH in advancing biomarker research and understanding clinical heterogeneity in neuropsychiatric disorders like ASD.