Higher-order organization of multivariate time series

Nature Physics, Год журнала: 2023, Номер unknown

Опубликована: Янв. 2, 2023

Язык: Английский

---

Deep learning models reveal replicable, generalizable, and behaviorally relevant sex differences in human functional brain organization

Proceedings of the National Academy of Sciences, Год журнала: 2024, Номер 121(9)

Опубликована: Фев. 20, 2024

Sex plays a crucial role in human brain development, aging, and the manifestation of psychiatric neurological disorders. However, our understanding sex differences functional organization their behavioral consequences has been hindered by inconsistent findings lack replication. Here, we address these challenges using spatiotemporal deep neural network (stDNN) model to uncover latent dynamics that distinguish male female brains. Our stDNN accurately differentiated brains, demonstrating consistently high cross-validation accuracy (>90%), replicability, generalizability across multisession data from same individuals three independent cohorts (N ~ 1,500 young adults aged 20 35). Explainable AI (XAI) analysis revealed features associated with default mode network, striatum, limbic exhibited significant (effect sizes > 1.5) sessions cohorts. Furthermore, XAI-derived predicted sex-specific cognitive profiles, finding was also independently replicated. results demonstrate are not only highly replicable generalizable but behaviorally relevant, challenging notion continuum male-female organization. underscore as biological determinant organization, have implications for developing personalized biomarkers disorders, provide innovative AI-based computational tools future research.

Язык: Английский

---

Brain structure-function coupling provides signatures for task decoding and individual fingerprinting

NeuroImage, Год журнала: 2022, Номер 250, С. 118970 - 118970

Опубликована: Фев. 4, 2022

Brain signatures of functional activity have shown promising results in both decoding brain states, meaning distinguishing between different tasks, and fingerprinting, that is identifying individuals within a large group. Importantly, these do not account for the underlying anatomy on which function takes place. Structure-function coupling based graph signal processing (GSP) has recently revealed meaningful spatial gradient from unimodal to transmodal regions, average healthy subjects during resting-state. Here, we explore specificity structure-function distinct states (tasks) individual subjects. We used multimodal magnetic resonance imaging 100 unrelated Human Connectome Project rest seven tasks adopted support vector machine classification approach with various cross-validation settings. found measures allow accurate classifications task fingerprinting. In particular, key information fingerprinting more liberal portion signals, contributions strikingly localized fronto-parietal network. Moreover, signals showed strong correlation cognitive traits, assessed partial least square analysis, corroborating its relevance By introducing new perspective GSP-based filtering FC decomposition, show provides class cognition organization at tasks. Further, they provide insights clarifying role low high frequencies structural connectome, leading understanding where characterizing can be across connectome spectrum.

Язык: Английский

---

The INSIDEOUT framework provides precise signatures of the balance of intrinsic and extrinsic dynamics in brain states

Communications Biology, Год журнала: 2022, Номер 5(1)

Опубликована: Июнь 10, 2022

Abstract Finding precise signatures of different brain states is a central, unsolved question in neuroscience. We reformulated the problem to quantify ‘inside out’ balance intrinsic and extrinsic dynamics states. The difference state can be described as differences detailed causal interactions found underlying dynamics. used thermodynamics framework breaking captured by level asymmetry temporal processing, i.e. arrow time. Specifically, was computed time-shifted correlation matrices for forward reversed time series, reflecting non-reversibility/non-equilibrium. precise, distinguishing terms reversibility hierarchy large-scale three radically (awake, deep sleep anaesthesia) electrocorticography data from non-human primates. Significantly lower levels were anaesthesia compared wakefulness. Non-wakeful also showed flatter hierarchy, diversity across brain. Overall, this provides states, perhaps conscious awareness.

Язык: Английский

---

Identifying canonical and replicable multi‐scale intrinsic connectivity networks in 100k+ resting‐state fMRI datasets

Human Brain Mapping, Год журнала: 2023, Номер 44(17), С. 5729 - 5748

Опубликована: Окт. 3, 2023

Abstract Despite the known benefits of data‐driven approaches, lack approaches for identifying functional neuroimaging patterns that capture both individual variations and inter‐subject correspondence limits clinical utility rsfMRI its application to single‐subject analyses. Here, using data from over 100k individuals across private public datasets, we identify replicable multi‐spatial‐scale canonical intrinsic connectivity network (ICN) templates via use multi‐model‐order independent component analysis (ICA). We also study feasibility estimating subject‐specific ICNs spatially constrained ICA. The results show subject‐level ICN estimations vary as a function itself, length, spatial resolution. In general, large‐scale require less achieve specific levels (within‐ between‐subject) similarity with their templates. Importantly, increasing length can reduce an ICN's specificity, suggesting longer scans may not always be desirable. find positive linear relationship between smoothness (possibly due averaging dynamics), studies examining optimized should consider smoothness. Finally, consistency in estimated full subsets different lengths suggests lower within‐subject shorter is wholly defined by reliability estimates, but indication meaningful brain dynamics which average out increases.

Язык: Английский

---

The neural oscillatory mechanism underlying human brain fingerprint recognition using a portable EEG acquisition device

NeuroImage, Год журнала: 2024, Номер 294, С. 120637 - 120637

Опубликована: Май 5, 2024

In recent years, brainprint recognition has emerged as a novel method of personal identity verification. Although studies have demonstrated the feasibility this technology, some limitations hinder its further development into society, such insufficient efficiency (extended wear time for multi-channel EEG cap), complex experimental paradigms (more in learning and completing experiments), unclear neurobiological characteristics (lack intuitive biomarkers an inability to eliminate impact noise on individual differences). Overall, these are due incomplete understanding underlying neural mechanisms. Therefore, study aims investigate mechanisms behind brainwave simplify operation process. We recorded prefrontal resting-state data from 40 participants, which is followed up over nine months using single-channel portable device. found that devices can effectively stably capture different subjects alpha band (8-13Hz) long periods, well capturing their differences (no peak, 1 or 2 peaks). Through correlation analysis, alpha-band activity reveal uniqueness compared others within one minute. used descriptive model dissect oscillatory non-oscillatory components band, demonstrating contributions fine features (especially amplitude bandwidth). Our validated oscillation The various oscillations will contribute accuracy recognition, providing new insights future technology.

Язык: Английский

---

Dynamic sensitivity analysis: Defining personalised strategies to drive brain state transitions via whole brain modelling

Computational and Structural Biotechnology Journal, Год журнала: 2022, Номер 21, С. 335 - 345

Опубликована: Дек. 1, 2022

Traditionally, in neuroimaging, model-free analyses are used to find significant differences between brain states via signal detection theory. Depending on the a priori assumptions about underlying data, different spatio-temporal features can be analysed. Alternatively, model-based techniques infer from data and compare significance model parameters. However, assess transitions one state another remains challenge current paradigms. Here, we introduce "Dynamic Sensitivity Analysis" framework that quantifies terms of stimulation ability rebalance activity towards target such as healthy dynamics. In practice, it means building whole-brain fitted description dynamics, applying systematic stimulations in-silico optimal strategy drive dynamics state. Further, show how Dynamic Analysis extends various paradigms, ultimately contributing improving efficacy personalised clinical interventions.

Язык: Английский

---

Evidence for increased parallel information transmission in human brain networks compared to macaques and male mice

Nature Communications, Год журнала: 2023, Номер 14(1)

Опубликована: Дек. 11, 2023

Abstract Brain communication, defined as information transmission through white-matter connections, is at the foundation of brain’s computational capacities that subtend almost all aspects behavior: from sensory perception shared across mammalian species, to complex cognitive functions in humans. How did communication strategies macroscale brain networks adapt evolution accomplish increasingly functions? By applying a graph- and information-theory approach assess information-related pathways male mouse, macaque human brains, we show gap between selective non-human mammals, where regions share single polysynaptic pathways, parallel humans, multiple pathways. In acts major connector unimodal transmodal systems. The layout unique individuals different pointing individual-level specificity routing architecture. Our work provides evidence patterns are tied networks.

Язык: Английский

---

A Review of Fingerprint Sensors: Mechanism, Characteristics, and Applications

Micromachines, Год журнала: 2023, Номер 14(6), С. 1253 - 1253

Опубликована: Июнь 14, 2023

Identification technology based on biometrics is a branch of research that employs the unique individual traits humans to authenticate identity, which most secure method identification its exceptional high dependability and stability human biometrics. Common biometric identifiers include fingerprints, irises, facial sounds, among others. In realm recognition, fingerprint recognition has gained success with convenient operation fast identif ication speed. Different collecting techniques, supply information for systems, have attracted significant deal interest in authentication regarding systems. This work presents several acquisition such as optical capacitive ultrasonic, analyzes types structures. addition, pros drawbacks various sensor types, well limits benefits optical, capacitive, ultrasonic kinds, are discussed. It necessary stage application Internet Things (IoT).

Язык: Английский

---

Higher-order connectomics of human brain function reveals local topological signatures of task decoding, individual identification, and behavior

Nature Communications, Год журнала: 2024, Номер 15(1)

Опубликована: Ноя. 26, 2024

Abstract Traditional models of human brain activity often represent it as a network pairwise interactions between regions. Going beyond this limitation, recent approaches have been proposed to infer higher-order from temporal signals involving three or more However, day remains unclear whether methods based on inferred outperform traditional ones for the analysis fMRI data. To address question, we conducted comprehensive using time series 100 unrelated subjects Human Connectome Project. We show that greatly enhance our ability decode dynamically various tasks, improve individual identification unimodal and transmodal functional subsystems, strengthen significantly associations behavior. Overall, approach sheds new light organization series, improving characterization dynamic group dependencies in rest revealing vast space unexplored structures within data, which may remain hidden when approaches.

Язык: Английский

---