Thursday, Jun 26: 11:30 AM - 12:45 PM
Oral Sessions
Brisbane Convention & Exhibition Centre
Room: Great Hall
Presentations
Current efforts on the prediction of cognitive decline from demographic, genetic, and brain imaging features primarily focus on: (1) predicting future diagnoses1, (2) generating point estimates (e.g., expected values)2, and (3) using fixed time windows3,4. A probabilistic approach predicting future cognitive decline trajectories offers significant advantages by capturing the uncertainty of predictions, accommodating arbitrary time intervals, and enabling personalized trajectories that account for individual variability in disease progression. Despite benefits, evaluating probabilistic forecasts poses greater challenges than point estimates due to the need for robust calibration and discrimination assessments, ensuring that predicted probabilities can be used for adjustable decision thresholds with confidence guarantees that meet requirements for future translation of such models into actionable insights5. In this work, we introduce probabilistic forecasting of future Clinical Dementia Rating Sum of Boxes (CDR-SOB)6, often used as primary outcome measures in clinical trials of Alzheimer's disease (AD), and the multidimensional evaluation of performance.
Large-scale MRI datasets have established brain-wide association studies (BWAS) as crucial for mapping individual variability in brain function and behavior. While BWAS has primarily focused on inter-regional connectivity via resting-state functional connectivity (RSFC), intra-regional neural dynamics remain underexplored. Neural variability within regions provides critical insights into brain-behavior relationships, but existing metrics of resting-state regional dynamics (RSRD), such as variability, fluctuation, and correlation, offer fragmented perspectives, risking oversimplification. Data-driven approaches integrating diverse temporal features are essential to fully realize RSRD's potential. Incorporating dynamic elements creates nuanced, individual-specific profiles of brain activity, offering deeper insights into behavior across spatial and temporal scales (Petersen et al., 2024). Using the hctsa toolbox (Fulcher & Jones, 2017), we analyzed three independent lifespan datasets (N=30,138; ages 8–82) and developed RSRD profiles capturing multifaceted temporal patterns of regional brain activity. Our goals (Fig. 1A) were to construct robust profiles, identify behavior-specific features, and evaluate their generalizability across populations and life stages.
Presenter
Xiaohan Tian, beijing normal university
beijing normal university
beijing, beijing
China
Diffusion MRI (dMRI) is sensitive to small changes in brain microstructure and may offer sensitivity to early Alzheimer's disease (AD) neuropathology that precedes macrostructural brain changes. Subtle pathology associated with early amyloid (Aβ) deposition may be better captured by dMRI measures in cortical gray matter (GM), where the earliest AD histopathological changes occur, compared to GM volume or thickness, the most commonly used MRI biomarkers. Recently, single-shell adaptations of multi-shell NODDI [1] and mean apparent propagator (MAP)-MRI [2] dMRI models have been proposed: NODDI-DTI [3] and MAP-AMURA [4], respectively. These models may mimic the sensitivity of multi-shell models to sources of non-Gaussianity in the GM, including dispersion and restriction, and may help to identify early correlates of Aβ accumulation before neurodegeneration and cognitive impairment occur. Here, we evaluated relationships between Aβ PET and advanced single-shell cortical microstructural measures in cognitively normal (CN) individuals from 3 AD studies. For comparison, more conventional DTI and cortical thickness (CTh) measures were also evaluated.
Presenter
Talia Nir, Keck School of Medicine, University of Southern California
Mark & Mary Stevens Neuroimaging & Informatics Institute,
Marina Del Rey, CA
United States
Individual differences in brain function have been prominently observed during development. To characterize this uniqueness, researchers use functional magnetic resonance imaging (fMRI) based brain fingerprinting, achieving high-precision individual identification in adults (Finn, 2015). Current methods demonstrate great potential in capturing functional fingerprints, yet more refined methodologies are needed to delineate the unique developmental trajectory of the brain. Here, we propose a deep learning framework for longitudinal functional fingerprinting in early adolescents, identifying unique brain functional patterns relating to cognitive abilities and genetics.
Presenter
Rui Xu, Beijing University of Posts and Telecommunications
Beijing University of Posts and Telecommunications
Beijing, Beijing
China
Biological aging, in both health and disease, is linked to changes in brain architecture. Neuroimaging has been used extensively to study the impact of aging on brain structure, revealing reductions in cortical volume, surface area, thickness, and gyrification. New evidence suggests that these structural changes reflect dysconnectivity within brain networks that are particularly vulnerable to aging [1].
Structural connectivity is often studied using tractography derived from diffusion-weighted MRI data [2]. Yet, this approach is sensitive to false-positive connections and does not leverage information sensitive to cytoarchitectural changes (i.e. T1w, T2w, T2*). Morphometric Inverse Divergence (MIND) has been proposed as an alternative strategy, providing a measure of within-subject architectonic similarity between cortical areas based on multivariate distributions of vertex-wise MRI data for macro and microstructural features between region pair [3].
In this study, we will evaluate the sensitivity of MIND in detecting lifespan changes in global brain architecture in a cohort of N=63 macaque monkeys.
Presenter
Melina Tsotras, New York University New York, NY
United States
Resting functional MRI (rsfMRI), which relies on the temporal synchrony of BOLD signals among different brain regions, has been widely used to characterize the maturation of canonical brain functional networks throughout early infancy (Gao, 2014). Recently, Luppi et al. advanced this field by proposing an information-resolved framework to decompose BOLD signals into synergistic and redundant neural information processing among brain regions (Luppi, 2022). They reported that redundant interactions (RI) are more associated with basic brain functional networks, indicating network robustness. In contrast, synergistic interactions (SI) are linked to higher-order functional networks, reflecting integration to meet complex cognitive demands. While their results provide invaluable new insights into human neurocognitive architecture, the emergence of SI and RI during early infancy remains largely unknown. This study aimed to explore the developmental trajectories of SI and RI from preterm to term infants' brains, providing insights into network-specific vulnerabilities and their implications for neurodevelopment.
Presenter
Xinjie Qian, University of North Carolina at Chapel Hill Chapel Hill, NC
United States