Precision Functional Mapping of Individual Human Brains

Neuron, Год журнала: 2017, Номер 95(4), С. 791 - 807.e7

Опубликована: Июль 27, 2017

Язык: Английский

---

What Is the Test-Retest Reliability of Common Task-Functional MRI Measures? New Empirical Evidence and a Meta-Analysis

Psychological Science, Год журнала: 2020, Номер 31(7), С. 792 - 806

Опубликована: Июнь 3, 2020

Identifying brain biomarkers of disease risk is a growing priority in neuroscience. The ability to identify meaningful limited by measurement reliability; unreliable measures are unsuitable for predicting clinical outcomes. Measuring activity using task functional MRI (fMRI) major focus biomarker development; however, the reliability fMRI has not been systematically evaluated. We present converging evidence demonstrating poor task-fMRI measures. First, meta-analysis 90 experiments ( N = 1,008) revealed overall reliability—mean intraclass correlation coefficient (ICC) .397. Second, test-retest reliabilities priori regions interest across 11 common tasks collected Human Connectome Project 45) and Dunedin Study 20) were (ICCs .067–.485). Collectively, these findings demonstrate that currently suitable discovery or individual-differences research. review how this state affairs came be highlight avenues improving reliability.

Язык: Английский

---

Establishment of Best Practices for Evidence for Prediction

JAMA Psychiatry, Год журнала: 2019, Номер 77(5), С. 534 - 534

Опубликована: Ноя. 27, 2019

Importance

Great interest exists in identifying methods to predict neuropsychiatric disease states and treatment outcomes from high-dimensional data, including neuroimaging genomics data. The goal of this review is highlight several potential problems that can arise studies aim establish prediction.

Observations

A number have claimed prediction while establishing only correlation, which an inappropriate use the statistical meaning Statistical associations do not necessarily imply ability make predictions a generalized manner; evidence for thus requires testing model on data separate those used estimate model's parameters. This article discusses various measures predictive performance limitations some commonly measures, with focus importance using multiple when assessing performance. For classification, area under receiver operating characteristic curve appropriate measure; regression analysis, correlation should be avoided, median absolute error preferred.

Conclusions Relevance

To ensure accurate estimates validity, recommended best practices modeling include following: (1) in-sample fit indices reported as accuracy, (2) cross-validation procedure encompass all operations applied (3) analyses performed samples smaller than hundred observations, (4) accuracy examined reported, (5) coefficient determination computed sums squares formulation coefficient, (6) k-fold rather leave-one-out used.

Язык: Английский

---

An open resource for transdiagnostic research in pediatric mental health and learning disorders

Scientific Data, Год журнала: 2017, Номер 4(1)

Опубликована: Дек. 19, 2017

Abstract Technological and methodological innovations are equipping researchers with unprecedented capabilities for detecting characterizing pathologic processes in the developing human brain. As a result, ambitions to achieve clinically useful tools assist diagnosis management of mental health learning disorders gaining momentum. To this end, it is critical accrue large-scale multimodal datasets that capture broad range commonly encountered clinical psychopathology. The Child Mind Institute has launched Healthy Brain Network (HBN), an ongoing initiative focused on creating sharing biobank data from 10,000 New York area participants (ages 5–21). HBN Biobank houses about psychiatric, behavioral, cognitive, lifestyle phenotypes, as well brain imaging (resting naturalistic viewing fMRI, diffusion MRI, morphometric MRI), electroencephalography, eye-tracking, voice video recordings, genetics actigraphy. Here, we present rationale, design implementation protocols. We describe first release ( n =664) potential advance related areas (e.g., biophysical modeling, analysis).

Язык: Английский

---

Task-induced brain state manipulation improves prediction of individual traits

Nature Communications, Год журнала: 2018, Номер 9(1)

Опубликована: Июль 12, 2018

Abstract Recent work has begun to relate individual differences in brain functional organization human behaviors and cognition, but the best state reveal such relationships remains an open question. In two large, independent data sets, we here show that cognitive tasks amplify trait-relevant patterns of connectivity, predictive models built from task fMRI outperform resting-state data. Further, certain consistently yield better predictions fluid intelligence than others, generates best-performing varies by sex. By considering task-induced sex, model explains over 20% variance scores, as compared <6% explained rest-based models. This suggests identifying inducing right a given group can brain-behavior relationships, motivating paradigm shift rest- task-based connectivity analyses.

Язык: Английский

---

Predicting brain-age from multimodal imaging data captures cognitive impairment

NeuroImage, Год журнала: 2016, Номер 148, С. 179 - 188

Опубликована: Ноя. 23, 2016

Язык: Английский

---

Studying individual differences in human adolescent brain development

Nature Neuroscience, Год журнала: 2018, Номер 21(3), С. 315 - 323

Опубликована: Янв. 31, 2018

Язык: Английский

---

Human brain mapping: A systematic comparison of parcellation methods for the human cerebral cortex

NeuroImage, Год журнала: 2017, Номер 170, С. 5 - 30

Опубликована: Апрель 13, 2017

Язык: Английский

---

Small sample sizes reduce the replicability of task-based fMRI studies

Communications Biology, Год журнала: 2018, Номер 1(1)

Опубликована: Май 30, 2018

Despite a growing body of research suggesting that task-based functional magnetic resonance imaging (fMRI) studies often suffer from lack statistical power due to too-small samples, the proliferation such underpowered continues unabated. Using large independent samples across eleven tasks, we demonstrate impact sample size on replicability, assessed at different levels analysis relevant fMRI researchers. We find degree replicability for typical sizes is modest and much larger than (e.g., N = 100) produce results fall well short perfectly replicable. Thus, our join existing line work advocating sizes. Moreover, because test over fairly range use intuitive metrics hope are more understandable convincing researchers who may have found previous inaccessible.

Язык: Английский

---

Can brain state be manipulated to emphasize individual differences in functional connectivity?

NeuroImage, Год журнала: 2017, Номер 160, С. 140 - 151

Опубликована: Март 31, 2017

While neuroimaging studies typically collapse data from many subjects, brain functional organization varies between individuals, and characterizing this variability is crucial for relating activity to behavioral phenotypes. Rest has become the default state probing individual differences, chiefly because it easy acquire a supposed neutral backdrop. However, assumption that rest optimal condition differences research largely untested. In fact, other states may afford better ratio of within- between-subject variability, facilitating biomarker discovery. Depending on trait or behavior under study, certain tasks bring out meaningful idiosyncrasies across essentially enhancing signal in networks interest beyond what can be measured at rest. Here, we review theoretical considerations existing work how influences connectivity, present some preliminary analyses conditions using Human Connectome Project, outline questions future study.

Язык: Английский

---