A Radiomic Approach to Clinical MRI Refines the Thalamus-Cognition Link in Multiple Sclerosis

Area of Science:

• Neuroimaging
• Radiology
• Cognitive Neuroscience

Background:

• Multiple sclerosis (MS) significantly impacts cognitive function.
• Conventional volumetric analyses may not capture subtle brain changes in MS.
• Radiomics offers a method to extract detailed imaging features beyond simple volume measurements.

Purpose of the Study:

• To investigate the association between radiomic features (RF) of thalamic nuclei and cognitive performance in individuals with MS.
• To compare the efficacy of radiomic features against volumetric measures in predicting cognitive outcomes.
• To explore the utility of radiomics in understanding MS-related cognitive deficits.

Main Methods:

• Radiomic features and volumes were extracted from the thalamus and its nuclei in 601 participants (342 people with MS, 259 healthy controls).
• Cognitive performance was assessed using standardized tests (SDMT, PASAT, DSST).
• Multivariate regression and LASSO models were employed to identify and validate RF associated with cognitive scores, comparing them with volumetric data.

Main Results:

• Twenty-eight radiomic feature-region of interest pairs were significantly associated with cognitive performance in people with MS.
• These radiomic features demonstrated stronger associations with cognitive tests (SDMT, PASAT) than volumetric measures.
• Radiomic features from the putamen showed limited association with cognitive performance compared to thalamic features.

Conclusions:

• Thalamic radiomic features are strongly linked to cognitive performance in people with MS.
• Radiomics surpasses volumetric analysis in identifying imaging biomarkers for cognitive impairment in MS.
• Radiomic analysis holds promise for developing sensitive biomarkers for MS-related cognitive dysfunction.