Detecting frontotemporal dementia syndromes using MRI biomarkers

Area of Science:

• Neuroimaging
• Biomarker Discovery
• Dementia Diagnostics

Background:

• Diagnosing frontotemporal dementia (FTD) presents significant clinical challenges.
• Magnetic resonance imaging (MRI) analysis of regional brain atrophy patterns offers potential for improved diagnostic assessment.
• Automated imaging biomarkers could enhance the differentiation of FTD subtypes.

Purpose of the Study:

• To develop and validate automated MRI-based imaging biomarkers for differentiating FTD subtypes.
• To distinguish FTD from other dementia diagnoses and from each other.
• To assess the diagnostic performance of novel atrophy pattern indices.

Main Methods:

• Retrospective analysis of 1213 patients from memory clinic cohorts, including FTD, Alzheimer's disease, and other dementias.
• Derivation of three MRI atrophy biomarkers: anterior vs. posterior index, asymmetry index, and temporal pole left index.
• Performance evaluation using AUC, sensitivity, and specificity, with a target specificity of 95% for FTD detection, validated in an independent cohort.

Main Results:

• The anterior vs. posterior index achieved 83% AUC for differentiating FTD from other groups (95% specificity).
• The asymmetry index excelled at separating primary progressive aphasia and behavioral variant FTD (85% AUC).
• The temporal pole left index showed high performance for semantic variant primary progressive aphasia detection (85% AUC).

Conclusions:

• Three quantitative MRI biomarkers demonstrate potential for enhancing FTD diagnostic assessment.
• These biomarkers can assist clinicians in differentiating FTD subtypes and other neurodegenerative conditions.
• Automated imaging analysis provides valuable supplementary data for complex dementia diagnoses.