Linking brain structure, cognition, and sleep: insights from clinical data.

Area of Science:

• Neuroscience
• Medical Imaging
• Sleep Medicine

Background:

• Routinely collected clinical data, including brain MRI and polysomnography (PSG), often contain valuable information.
• Understanding the interplay between brain structure, physiology, and cognition is crucial for diagnosing and managing neurological and psychiatric conditions.

Purpose of the Study:

• To investigate relationships between brain physiology, structure, and cognition using noisy, routinely collected clinical data.
• To explore both hypothesis-driven and data-driven associations between sleep (PSG), brain structure (MRI), and cognitive function.

Main Methods:

• Analysis of data from 623 patients with sleep PSG and brain MRI data, including 160 with cognitive evaluations.
• Utilized SynthSeg for high-quality brain MRI segmentation and explored associations between PSG, MRI, and cognitive scores.
• Employed both a priori hypotheses and exploratory data-driven approaches to examine age-structure-physiology-cognition links.

Main Results:

• Positive correlations found between PSG delta power and frontal/thalamic volumes, and sleep spindle density and thalamic volume.
• Rapid eye movement (REM) duration and amygdala volume were positively associated with cognitive performance.
• Combined PSG and MRI features predicted chronological age (R2 = 0.67) and cognition (R2 = 0.40), with significant brain structure differences noted in dementia patients.

Conclusions:

• Routine clinical data, including brain MRI and sleep recordings, offers significant value for understanding brain-sleep-cognition relationships.
• These findings support the clinical utility of integrating physiological and structural brain data for improved patient assessment and management.