A self-administered, artificial intelligence (AI) platform for cognitive assessment in multiple sclerosis (MS).

Area of Science:

• Neuroscience
• Digital Health
• Clinical Assessment

Background:

• Cognitive impairment is a prevalent issue in multiple sclerosis (MS).
• Developing reliable cognitive measures is crucial for tracking disease progression and treatment efficacy in MS.
• Digital tools offer potential for accessible and repeatable cognitive assessments.

Purpose of the Study:

• To develop and validate a novel, brief, self-administered computerized cognitive test for MS patients.
• To assess the Integrated Cognitive Assessment's (ICA) reliability, validity, and diagnostic accuracy.
• To explore the association between ICA scores, neurofilament light (NfL) levels, and rehabilitation in MS.

Main Methods:

• Developed the 5-minute Integrated Cognitive Assessment (ICA) test, which is self-administered and language-independent.
• Conducted Substudy 1 with 91 MS patients and 83 healthy controls (HC) comparing ICA with BICAMS.
• In Substudy 2, analyzed ICA and SDMT scores in relation to serum NfL levels pre- and post-rehabilitation in 48 MS patients.

Main Results:

• The ICA demonstrated excellent test-retest reliability (r=0.94) and convergent validity with BICAMS.
• ICA accurately discriminated MS patients from HC (AUC=95%) and showed no learning bias.
• A strong negative association (r=-0.79) was found between ICA scores and serum NfL levels in MS patients.

Conclusions:

• The Integrated Cognitive Assessment (ICA) shows promise as a digital biomarker for cognitive impairment in MS.
• ICA's brevity, language independence, and lack of learning bias make it advantageous over standard tools.
• The ICA can potentially monitor patient response to therapeutic interventions and disease activity.