A Field Theory of Human Intelligence DOI Open Access

Alan Griswold

Published: Oct. 22, 2023

The standard model of human intelligence is a brain-centric and brain-specific depiction intelligence, it enjoys nearly universal acceptance within the research community. Nonetheless, does face some serious challenges, including lack specificity an inability to account for Flynn effect (other than assume that must be temporary aberration). What being presented here alternative one locates not brain but instead growing amount artificial structure contained environment. Although this field theory approach runs counter widely accepted model, offer several advantages. One, eschews any extraordinary biological or evolutionary assumptions regarding functioning brain. Two, provides specific observable description material intelligence. And three, gives straightforward elegant explanation effect. For these reasons, merits consideration.

Language: Английский

The multiverse of data preprocessing and analysis in graph-based fMRI: A systematic literature review of analytical choices fed into a decision support tool for informed analysis DOI Creative Commons
Daniel Kristanto,

Micha Burkhardt,

Christiane M. Thiel

et al.

Neuroscience & Biobehavioral Reviews, Journal Year: 2024, Volume and Issue: 165, P. 105846 - 105846

Published: Aug. 6, 2024

The large number of different analytical choices used by researchers is partly responsible for the challenge replication in neuroimaging studies. For an exhaustive robustness analysis, knowledge full space options essential. We conducted a systematic literature review to identify decisions functional data preprocessing and analysis emerging field cognitive network neuroscience. found 61 steps, with 17 them having debatable parameter choices. Scrubbing, global signal regression, spatial smoothing are among controversial steps. There no standardized order which steps applied, settings within several vary widely across By aggregating pipelines studies, we propose three taxonomic levels categorize choices: 1) inclusion or exclusion specific 2) tuning 3) distinct sequencing have developed decision support application high educational value called METEOR facilitate access design well-informed (multiverse) analysis.

Language: Английский

Citations

3
The multiverse of data preprocessing and analysis in graph-based fMRI: A systematic literature review of analytical choices fed into a decision support tool for informed analysis DOI Creative Commons
Daniel Kristanto,

Micha Burkhardt,

Christiane M. Thiel

et al.

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

Published: Jan. 15, 2024

Abstract The large number of different analytical choices researchers use may be partly responsible for the replication challenge in neuroimaging studies. For robustness analysis, knowledge full space options is essential. We conducted a systematic literature review to identify decisions functional data preprocessing and analysis emerging field cognitive network neuroscience. found 61 steps, with 17 them having debatable options. Scrubbing, global signal regression, spatial smoothing are among controversial steps. There no standardized order which steps applied, within several vary widely across By aggregating pipelines studies, we propose three taxonomic levels categorize choices: 1) inclusion or exclusion specific 2) distinct sequencing 3) parameter tuning To facilitate access data, developed decision support app high educational value called METEOR, allows explore as reference well-informed (multiverse) analysis. Highlights Data variability hinders replicability. Analysis multiple defensible examines results. identified 102 performing graph-fMRI Interactive visualization these available Shiny app.

Language: Английский

Citations

2
An Extended Active Learning Approach to Multiverse Analysis: Predictions of Latent Variables from Graph Theory Measures of the Human Connectome and Their Direct Replication DOI Open Access
Daniel Kristanto, Carsten Gießing, Merle Johanna Marek

et al.

Brainiacs Journal of Brain Imaging And Computing Sciences, Journal Year: 2023, Volume and Issue: 4(2)

Published: Dec. 23, 2023

Multiverse analysis has been proposed as a powerful technique to disclose the large number of degrees freedom in data preprocessing and that strongly contribute current replication crisis science.However, field imaging neuroscience, where multidimensional, complex noisy are measured, multiverse may be computationally infeasible.The possible forking paths given by different methodological decisions analytical choices is immense.Recently, Dafflon et al. (2022) an active learning approach alternative exhaustively exploring all paths.Here, we aimed extend their pipeline integrating latent underlying variables which not directly observable.The extension outcomes particularly valuable for computational psychiatry neurocognitive psychology, traits conceptualized common cause variety observable neural behavioral symptoms.To illustrate our test its direct replicability, analyzed individual organization topology functional brain networks two relatively samples from ABCD study dataset (N = 1491) HCP 833).Graph-theoretical parameters take into account both brain-wide region-specific network properties were used predictors variable reflecting general cognition.Our results demonstrate ability extended method selectively explore when predicting variable.First, low-dimensional space created with was able cluster according similarity.Second, successfully estimated prediction performance pipelines datasets.To interactively results, developed Shiny app visualize predictive accuracy resulting each path similarity between combinations processing choice.The code available at Github repository ExtendedAL.

Language: Английский

Citations

1
Unveiling the Core Functional Networks of Cognition: An Ontology-Guided Machine Learning Approach DOI Creative Commons
Guowei Wu, Zaixu Cui, Xiuyi Wang

et al.

NeuroImage, Journal Year: 2024, Volume and Issue: 298, P. 120804 - 120804

Published: Aug. 23, 2024

Deciphering the functional architecture that underpins diverse cognitive functions is fundamental quest in neuroscience. In this study, we employed an innovative machine learning framework integrated ontology with connectivity analysis to identify brain networks essential for cognition. We identified a core assembly of connectomes, primarily located within association cortex, which showed superior predictive performance compared two conventional methods widely previous research across various domains. Our approach achieved mean prediction accuracy 0.13 16 tasks, including working memory, reading comprehension, and sustained attention, outperforming traditional methods' 0.08. contrast, our method limited power sensory, motor, emotional functions, 0.03 9 relevant slightly lower than 0.04. These connectomes were further characterized by distinctive patterns resting-state connectivity, structural via white matter tracts, gene expression, highlighting their neurogenetic underpinnings. findings reveal domain-general network fingerprint pivotal cognition, offering novel computational explore neural foundations abilities.

Language: Английский

Citations

0
A Field Theory of Human Intelligence DOI Open Access

Alan Griswold

Published: Oct. 22, 2023

The standard model of human intelligence is a brain-centric and brain-specific depiction intelligence, it enjoys nearly universal acceptance within the research community. Nonetheless, does face some serious challenges, including lack specificity an inability to account for Flynn effect (other than assume that must be temporary aberration). What being presented here alternative one locates not brain but instead growing amount artificial structure contained environment. Although this field theory approach runs counter widely accepted model, offer several advantages. One, eschews any extraordinary biological or evolutionary assumptions regarding functioning brain. Two, provides specific observable description material intelligence. And three, gives straightforward elegant explanation effect. For these reasons, merits consideration.

Language: Английский

Citations

0