Developing a disease staging biomarker based on micro density patterns in bvFTD patients

Brisbane Convention & Exhibition Centre 

Frontotemporal dementia (FTD) is a neurodegenerative disorder affecting the frontal and temporal lobes (Grossman, 2023), with behavioral variant FTD (bvFTD) as the most common clinical presentation, characterized by changes in social behavior, personality, and executive function (Matarrubia, 2014). Staging dementia is critical for effective clinical management, as it helps tailor personalized care, and it provides a framework for documenting the impact of therapeutic interventions that may alter the course of the underlying disorder.
T1-weighted MRI is an important modality for diagnosis and clinical work-up, where visual inspection of cortical brain volume is part of the diagnostic criteria. Automated brain MRI segmentation methods are considered better than visual inspection alone, as they offer quantitative and repeatable measures of anatomical changes. However, volumetric analysis of structural MRI is limited to characterizing volume loss which is thought to occur later during the disease process. In contrast, texture analysis, which quantifies microstructural changes by examining relationships between signal intensities of neighboring voxels (Larroza, 2016), may detect physiological changes leading to neuronal loss.
We hypothesize that gray matter microstructural damage, quantified through texture analysis, can differentiate disease stages in bvFTD. Specifically, we focus on the sum average texture feature, which captures the relationship between radiolucent (dark, volume loss) and radiopaque (bright, dense) regions in MRI images. Neuron loss typically manifests as dark regions on T1-weighted MRI, with early changes often too subtle for visual detection. This study explores whether sum average can distinguish between mild and moderate dementia in bvFTD patients.