Discovering causal relations and equations from data

Physics Reports, Год журнала: 2023, Номер 1044, С. 1 - 68

Опубликована: Ноя. 7, 2023

Physics is a field of science that has traditionally used the scientific method to answer questions about why natural phenomena occur and make testable models explain phenomena. Discovering equations, laws, principles are invariant, robust, causal been fundamental in physical sciences throughout centuries. Discoveries emerge from observing world and, when possible, performing interventions on system under study. With advent big data data-driven methods, fields equation discovery have developed accelerated progress computer science, physics, statistics, philosophy, many applied fields. This paper reviews concepts, relevant works broad physics outlines most important challenges promising future lines research. We also provide taxonomy for discovery, point out connections, showcase comprehensive case studies Earth climate sciences, fluid dynamics mechanics, neurosciences. review demonstrates discovering laws relations by revolutionised with efficient exploitation observational simulations, modern machine learning algorithms combination domain knowledge. Exciting times ahead opportunities improve our understanding complex systems.

Язык: Английский

---

Brain clocks capture diversity and disparities in aging and dementia across geographically diverse populations

Nature Medicine, Год журнала: 2024, Номер 30(12), С. 3646 - 3657

Опубликована: Авг. 26, 2024

Abstract Brain clocks, which quantify discrepancies between brain age and chronological age, hold promise for understanding health disease. However, the impact of diversity (including geographical, socioeconomic, sociodemographic, sex neurodegeneration) on brain-age gap is unknown. We analyzed datasets from 5,306 participants across 15 countries (7 Latin American Caribbean (LAC) 8 non-LAC countries). Based higher-order interactions, we developed a deep learning architecture functional magnetic resonance imaging (2,953) electroencephalography (2,353). The comprised healthy controls individuals with mild cognitive impairment, Alzheimer disease behavioral variant frontotemporal dementia. LAC models evidenced older ages (functional imaging: mean directional error = 5.60, root square (r.m.s.e.) 11.91; electroencephalography: 5.34, r.m.s.e. 9.82) associated frontoposterior networks compared models. Structural socioeconomic inequality, pollution disparities were influential predictors increased gaps, especially in ( R ² 0.37, F 0.59, 6.9). An ascending to impairment was found. In LAC, observed larger gaps females control groups respective males. results not explained by variations signal quality, demographics or acquisition methods. These findings provide quantitative framework capturing accelerated aging.

Язык: Английский

---

A synergetic turn in cognitive neuroscience of brain diseases

Trends in Cognitive Sciences, Год журнала: 2024, Номер 28(4), С. 319 - 338

Опубликована: Янв. 21, 2024

Язык: Английский

---

Source space connectomics of neurodegeneration: One-metric approach does not fit all

Neurobiology of Disease, Год журнала: 2023, Номер 179, С. 106047 - 106047

Опубликована: Фев. 23, 2023

Brain functional connectivity in dementia has been assessed with dissimilar EEG metrics and estimation procedures, thereby increasing results' heterogeneity. In this scenario, joint analyses integrating information from different may allow for a more comprehensive characterization of brain interactions subtypes. To test hypothesis, resting-state electroencephalogram (rsEEG) was recorded individuals Alzheimer's Disease (AD), behavioral variant frontotemporal (bvFTD), healthy controls (HCs). Whole-brain estimated the source space using 101 types connectivity, capturing linear nonlinear both time frequency-domains. Multivariate machine learning progressive feature elimination run to discriminate AD HCs, bvFTD based on i) frequency bands, ii) complementary frequency-domain (e.g., instantaneous, lagged, total connectivity), iii) time-domain linearity assumption Pearson correlation coefficient mutual information). <10% all possible connections were responsible differences between patients controls, atypical never captured by >1/4 measures. Joint revealed patterns hypoconnectivity (patientsHCs) groups mainly identified regions. These atypicalities differently frequency- metrics, bandwidth-specific fashion. The multi-metric representation whole-brain evidenced inadequacy single-metric approaches, resulted valid alternative selection problem connectivity. reveal interdependence that are overlooked single contributing reliable interpretable description neurodegeneration.

Язык: Английский

---

Model-based whole-brain perturbational landscape of neurodegenerative diseases

eLife, Год журнала: 2023, Номер 12

Опубликована: Март 30, 2023

The treatment of neurodegenerative diseases is hindered by lack interventions capable steering multimodal whole-brain dynamics towards patterns indicative preserved brain health. To address this problem, we combined deep learning with a model reproducing functional connectivity in patients diagnosed Alzheimer’s disease (AD) and behavioral variant frontotemporal dementia (bvFTD). These models included disease-specific atrophy maps as priors to modulate local parameters, revealing increased stability hippocampal insular signatures AD bvFTD, respectively. Using variational autoencoders, visualized different pathologies their severity the evolution trajectories low-dimensional latent space. Finally, perturbed reveal key AD- bvFTD-specific regions induce transitions from pathological healthy states. Overall, obtained novel insights on progression control means external stimulation, while identifying dynamical mechanisms that underlie alterations neurodegeneration.

Язык: Английский

---

Brain health in diverse settings: How age, demographics and cognition shape brain function

NeuroImage, Год журнала: 2024, Номер 295, С. 120636 - 120636

Опубликована: Май 21, 2024

Diversity in brain health is influenced by individual differences demographics and cognition. However, most studies on diseases have typically controlled for these factors rather than explored their potential to predict signals. Here, we assessed the role of (age, sex, education; n = 1,298) cognition (n 725) as predictors different metrics usually used case-control studies. These included power spectrum aperiodic (1/f slope, knee, offset) metrics, well complexity (fractal dimension estimation, permutation entropy, Wiener spectral structure variability) connectivity (graph-theoretic mutual information, conditional organizational information) from source space resting-state EEG activity a diverse sample global south north populations. Brain-phenotype models were computed using reflecting local (power components) dynamics interactions (complexity graph-theoretic measures). Electrophysiological modulated despite varied methods data acquisition assessments across multiple centers, indicating that results unlikely be accounted methodological discrepancies. Variations signals mainly age cognition, while education sex exhibited less importance. Power measures sensitive capturing differences. Older age, poorer being male associated with reduced alpha power, whereas older network integration segregation. Findings suggest basic impact core function are standard Considering variability diversity settings would contribute more tailored understanding function.

Язык: Английский

---

Alzheimer’s disease diagnosis using rhythmic power changes and phase differences: a low-density EEG study

Frontiers in Aging Neuroscience, Год журнала: 2025, Номер 16

Опубликована: Янв. 17, 2025

The future emergence of disease-modifying treatments for dementia highlights the urgent need to identify reliable and easily accessible tools diagnosing Alzheimer's disease (AD). Electroencephalography (EEG) is a non-invasive cost-effective technique commonly used in study neurodegenerative disorders. However, specific alterations EEG biomarkers associated with AD remain unclear when using limited number electrodes. We studied pathological characteristics low-density data collected from 26 29 healthy controls (HC) during both eye closed (EC) opened (EO) resting conditions. analysis including power spectrum, phase lock value (PLV), weighted lag index (wPLI) power-to-power frequency coupling (theta/beta) were applied extract features delta, theta, alpha, beta bands. During EC condition, group exhibited decreased alpha compared HC. Additionally, PLV wPLI theta band indicated that brain network predominantly involved frontal region opposite changes. Moreover, had increased central regions. Surprisingly, no difference was found EO condition. Notably, functional connectivity within fronto-central lobe EO. More importantly, combination quantitative improved inter-group classification accuracy support vector machine (SVM) older adults AD. These findings highlight complementary nature conditions assessing differentiating cohorts. Our results underscore potential utilizing resting-state paradigms, combined learning techniques, improve identification

Язык: Английский

---

Neuroimaging modalities in the detection of Alzheimer's disease-associated biomarkers

Deleted Journal, Год журнала: 2023, Номер 3

Опубликована: Янв. 1, 2023

Abstract Alzheimer's disease (AD) is the most common cause of dementia. Neuropathological changes in AD patients occur up to 10–20 years before emergence clinical symptoms. Specific diagnosis and appropriate intervention strategies are crucial during phase mild cognitive impairment (MCI) AD. The detection biomarkers has emerged as a promising tool for tracking efficacy potential therapies, making an early diagnosis, prejudging treatment prognosis. Specifically, multiple neuroimaging modalities, including magnetic resonance imaging (MRI), positron emission tomography, optical imaging, single photon emission-computed have provided few application. MRI modalities described this review include structural MRI, functional diffusion tensor spectroscopy, arterial spin labelling. These techniques allow presymptomatic diagnostic brains cognitively normal elderly people might also be used monitor progression after onset This highlights biomarkers, merits, demerits different their value MCI patients. Further studies necessary explore more overcome limitations inclusion criteria

Язык: Английский

---

The BrainLat project, a multimodal neuroimaging dataset of neurodegeneration from underrepresented backgrounds

Scientific Data, Год журнала: 2023, Номер 10(1)

Опубликована: Дек. 9, 2023

Abstract The Latin American Brain Health Institute (BrainLat) has released a unique multimodal neuroimaging dataset of 780 participants from American. includes 530 patients with neurodegenerative diseases such as Alzheimer’s disease (AD), behavioral variant frontotemporal dementia (bvFTD), multiple sclerosis (MS), Parkinson’s (PD), and 250 healthy controls (HCs). This (62.7 ± 9.5 years, age range 21–89 years) was collected through multicentric effort across five countries to address the need for affordable, scalable, available biomarkers in regions larger inequities. BrainLat is first regional collection clinical cognitive assessments, anatomical magnetic resonance imaging (MRI), resting-state functional MRI (fMRI), diffusion-weighted (DWI), high density electroencephalography (EEG) patients. In addition, it demographic information about harmonized recruitment assessment protocols. publicly encourage further research development tools health applications neurodegeneration based on neuroimaging, promoting variability inclusion underrepresented research.

Язык: Английский

---

Biomarkers of neurodegeneration across the Global South

The Lancet Healthy Longevity, Год журнала: 2024, Номер unknown, С. 100616 - 100616

Опубликована: Окт. 1, 2024

Язык: Английский

---