Multiscale brain modeling: bridging microscopic and macroscopic brain dynamics for clinical and technological applications

Frontiers in Cellular Neuroscience, Journal Year: 2025, Volume and Issue: 19

Published: Feb. 19, 2025

The brain's complex organization spans from molecular-level processes within neurons to large-scale networks, making it essential understand this multiscale structure uncover brain functions and address neurological disorders. Multiscale modeling has emerged as a transformative approach, integrating computational models, advanced imaging, big data bridge these levels of organization. This review explores the challenges opportunities in linking microscopic phenomena macroscopic functions, emphasizing methodologies driving progress field. It also highlights clinical potential including their role advancing artificial intelligence (AI) applications improving healthcare technologies. By examining current research proposing future directions for interdisciplinary collaboration, work demonstrates how can revolutionize both scientific understanding practice.

Language: Английский

---

Quality over quantity: powering neuroimaging samples in psychiatry

Neuropsychopharmacology, Journal Year: 2024, Volume and Issue: 50(1), P. 58 - 66

Published: June 20, 2024

Language: Английский

---

Toward a functional future for the cognitive neuroscience of human aging

Neuron, Journal Year: 2025, Volume and Issue: 113(1), P. 154 - 183

Published: Jan. 1, 2025

Language: Английский

---

Connectome-based fingerprinting: reproducibility, precision, and behavioral prediction

Neuropsychopharmacology, Journal Year: 2024, Volume and Issue: 50(1), P. 114 - 123

Published: Aug. 15, 2024

Language: Английский

---

External validation of machine learning models—registered models and adaptive sample splitting

GigaScience, Journal Year: 2025, Volume and Issue: 14

Published: Jan. 1, 2025

Abstract Background Multivariate predictive models play a crucial role in enhancing our understanding of complex biological systems and developing innovative, replicable tools for translational medical research. However, the complexity machine learning methods extensive data preprocessing feature engineering pipelines can lead to overfitting poor generalizability. An unbiased evaluation necessitates external validation, which involves testing finalized model on independent data. Despite its importance, validation is often neglected practice due associated costs. Results Here we propose that, maximal credibility, discovery should be separated by public disclosure (e.g., preregistration) processing steps weights. Furthermore, introduce novel approach optimize trade-off between efforts spent such studies. We show involving more than 3,000 participants from four different datasets any “sample size budget,” proposed adaptive splitting successfully identify optimal time stop so that performance maximized without risking low-powered, thus inconclusive, validation. Conclusion The design (implemented Python package “AdaptiveSplit”) may contribute addressing issues replicability, effect inflation, generalizability modeling

Language: Английский

---

Power and reproducibility in the external validation of brain-phenotype predictions

Nature Human Behaviour, Journal Year: 2024, Volume and Issue: 8(10), P. 2018 - 2033

Published: July 31, 2024

Language: Английский

---

Brain imaging studies of emotional well-being: a scoping review

Frontiers in Psychology, Journal Year: 2024, Volume and Issue: 14

Published: Jan. 5, 2024

This scoping review provides an overview of previous empirical studies that used brain imaging techniques to investigate the neural correlates emotional well-being (EWB). We compiled evidence on this topic into one accessible and usable document as a foundation for future research relationship between EWB brain. PRISMA 2020 guidelines were followed. located relevant articles by searching five electronic databases with 95 meeting our inclusion criteria. explored measures, modalities, designs, populations studied, approaches are currently in use characterize understand across literature. Of key concepts related EWB, vast majority investigated positive affect life satisfaction, followed sense meaning, goal pursuit, quality life. The functional MRI, EEG event-related potential-based study basis (predominantly experienced affect, affective perception, reward, emotion regulation). It is notable satisfaction have been studied significantly more often than other three aspects (i.e., life). Our findings suggest should diverse samples, especially children, individuals clinical disorders, from various geographic locations. Future directions theoretical implications discussed, including need longitudinal ecologically valid measures incorporate multi-level allowing researchers better evaluate relationships among behavioral, environmental, factors. Systematic registration https://osf.io/t9cf6/ .

Language: Английский

---

Replicability and generalizability in population psychiatric neuroimaging

Neuropsychopharmacology, Journal Year: 2024, Volume and Issue: 50(1), P. 52 - 57

Published: Aug. 30, 2024

Abstract Studies linking mental health with brain function in cross-sectional population-based association studies have historically relied on small, underpowered samples. Given the small effect sizes typical of such brain-wide associations, require samples into thousands to achieve statistical power necessary for replicability. Here, we detail how sample hampered replicability and provide size targets given established strength benchmarks. Critically, while will improve larger samples, it is not guaranteed that observed effects meaningfully apply target populations interest (i.e., be generalizable). We discuss important considerations related generalizability psychiatric neuroimaging an example failure due “shortcut learning” brain-based predictions phenotypes. Shortcut learning a phenomenon whereby machine models learn between unmeasured construct (the shortcut), rather than intended health. complex nature brain-behavior interactions, future epidemiological approaches large, diverse comprehensive assessment.

Language: Английский

---

Improving Predictability, Test-Retest Reliability and Generalisability of Brain-Wide Associations for Cognitive Abilities via Multimodal Stacking

bioRxiv (Cold Spring Harbor Laboratory), Journal Year: 2024, Volume and Issue: unknown

Published: May 5, 2024

Brain-wide association studies (BWASs) have attempted to relate cognitive abilities with brain phenotypes, but been challenged by issues such as predictability, test-retest reliability, and cross-cohort generalisability. To tackle these challenges, we proposed a machine-learning "stacking" approach that draws information from whole-brain magnetic resonance imaging (MRI) across different modalities, task-fMRI contrasts functional connectivity during tasks rest structural measures, into one prediction model. We benchmarked the benefits of stacking, using Human Connectome Projects: Young Adults (n=873, 22-35 years old) Projects-Aging (n=504, 35-100 Dunedin Multidisciplinary Health Development Study (Dunedin Study, n=754, 45 old). For stacked models led out-of-sample r ∼.5-.6 when predicting at time scanning, primarily driven contrasts. Notably, were able predict participants' ages 7, 9, 11 their multimodal MRI age 45, an 0.52. reached excellent level reliability (ICC>.75), even only together. generalisability, model non-task built dataset significantly predicted in other datasets. Altogether, stacking is viable undertake three challenges BWAS for abilities. Scientists had limited success MRI. machine learning method, called draw types Using large databases (n=2,131, 22-100 old), found make 1) closer actual scores applied new individual, not part modelling process, 2) reliable over times 3) applicable data collected groups scanners. Indeed, especially fMRI task contrasts, allowed us use people aged childhood reasonably well. Accordingly, may help realise its potential

Language: Английский

---

Multivariate analysis of multimodal brain structure predicts individual differences in risk and intertemporal preference

Published: July 8, 2024

Large changes to brain structure (e.g., from damage or disease) can explain alterations in behavior. It is therefore plausible that smaller structural differences healthy samples be used better understand and predict individual Despite the brain's multivariate distributed structure-to-function mapping, most studies have univariate analyses of measures. Here we a approach multimodal data set composed volumetric, surface-based, diffusion-based, functional resting-state MRI measures reliable risk intertemporal preferences. We show combining twelve led predictions across tasks than using any measure, by examining model coefficients, visualize relative contribution different regions. Using mapping combines many properties, along with reliably measured behavior phenotypes, may increase out-of-sample prediction accuracies insight into neural underpinnings. Furthermore, this methodological useful improve basic, translational, clinical research fields.

Language: Английский

---