Ventricular features as reliable differentiators between bvFTD and other dementias

Area of Science:

• Neuroimaging
• Neurology
• Artificial Intelligence

Background:

• Lateral ventricles are sensitive indicators of brain atrophy and disease progression in behavioral variant frontotemporal dementia (bvFTD).
• Accurate and early diagnosis of bvFTD is crucial for effective management and treatment.

Purpose of the Study:

• To investigate the utility of an automated tool using ventricular features for improving diagnostic accuracy in bvFTD.
• To assess the effectiveness of specific ventricular features in differentiating bvFTD from other neurodegenerative conditions.

Main Methods:

• Utilized data from 678 subjects, including bvFTD, other dementia cohorts, and normal controls, across 2750 timepoints.
• Segmented lateral ventricles and analyzed differences in ventricular features, such as antero-posterior ratio (APR), total ventricular volume, and left-right ratios.

Main Results:

• The ventricular antero-posterior ratio (APR) was significantly increased and showed faster progression in bvFTD compared to all other cohorts.
• Achieved 80% cross-validation accuracy (77% sensitivity, 82% specificity) in differentiating bvFTD using multiple ventricular features.
• Attained 76% accuracy using only the APR feature.

Conclusions:

• Ventricular features, especially APR, show potential as reliable and easily implementable diagnostic markers for bvFTD.
• The developed automated tool and model for ventricle feature estimation are publicly available for broader research application.