Building a Science of Individual Differences from fMRI

Dubois, Julien; Adolphs, Ralph

doi:10.1016/j.tics.2016.03.014

Cited by 570 publications

(489 citation statements)

References 201 publications

Supporting

Mentioning

482

Contrasting

Order By: Relevance

“…Relating behavioral assessments to brain images helps grounding inter-individual vari-5 ability in anatomical and functional aspects of brain organization (Dubois and Adolphs, 2016;Abi-Dargham and Horga, 2016). The corresponding imaging-based biomarkers can then serve as intermediate phenotypes, or neurophenotypes to characterize the subjects (Drysdale et al, 2016).…”

Section: Introductionmentioning

confidence: 99%

Joint prediction of multiple scores captures better individual traits from brain images

et al. 2017

View full text Add to dashboard Cite

To probe individual variations in brain organization, population imaging relates features of brain images to rich descriptions of the subjects such as genetic information or behavioral and clinical assessments. Capturing common trends across these measurements is important: they jointly characterize the disease status of patient groups. In particular, mapping imaging features to behavioral scores with predictive models opens the way toward more precise diagnosis. Here we propose to jointly predict all the dimensions (behavioral scores) that make up the individual profiles, using so-called multi-output models. This approach often boosts prediction accuracy by capturing latent shared information across scores. We demonstrate the efficiency of multi-output models on two independent resting-state fMRI datasets targeting different brain disorders (Alzheimer's Disease and schizophrenia). Furthermore, the model with joint prediction generalizes much better to a new cohort: a model learned on one study is more accurately transferred to an independent one. Finally, we show how multi-output models can easily be extended to multi-modal settings, combining heterogeneous data sources for a better overall accuracy.

show abstract

Section: Introductionmentioning

confidence: 99%

Joint prediction of multiple scores captures better individual traits from brain images

et al. 2017

View full text Add to dashboard Cite

show abstract

“…The most valuable comparisons would then be with two other datasets: with the same individual, but before the lesion (notoriously difficult to obtain since the lesions are generally unpredictable accidents of nature); and with other lesion patients who have damage in the same region, but don't show criminal behavior. Understanding the risks, mechanisms, and potential for brain-targeted treatments of criminal behavior will certainly require attention to single individuals, a trend now followed by much of neuroimaging (15).…”

Section: Reorganization and Individual Differencesmentioning

confidence: 99%

Searching for the neural causes of criminal behavior

Adolphs

Gläscher

Tranel

2017

Proc. Natl. Acad. Sci. U.S.A.

Self Cite

View full text Add to dashboard Cite

All behavior is proximally caused by the brain, but the neural causes of most complex behaviors are still not understood. Much of our ignorance stems from the fact that complex behavior depends on distributed neural control. Unlike a reflex, where the arc from sensation to action can be traced through a few synapses, most volitional behavior involves a dense causal web through which stimuli, memories, beliefs, and other factors exert their effects. Disruption anywhere in this causal web can produce effects that are difficult to trace back to their origin. Against this background, the finding that focal lesions of the ventromedial prefrontal cortex could lead to immoral and even criminal behavior generated considerable surprise and interest (1, 2). While a number of rare cases have now been described in whom a focal lesion caused criminality, these are neither very consistent (the lesions occur in several different anatomical locations) nor at all reliable (only a small fraction of patients, for any lesion location, show criminal behavior). To explain the effects of a lesion on criminal behavior, we need to understand what it is that the lesion does to the rest of the brain, a network-level understanding of lesion effects now provided by the new study of Darby et al. in PNAS (3). A Disconnection ApproachDarby et al. (3) began by searching the literature for cases documenting criminal behavior following a focal brain lesion. They found 17 of these, with lesions predominantly including the prefrontal cortex but also other regions. To estimate the distal effects of such brain damage, lesion network mapping (4) was used, showing which other brain regions, in healthy brains, would normally be functionally connected with the location of the lesion. This approach effectively treats the brain of the lesion patient as equivalent to a healthy brain in which all those regions functionally connected to the lesion site have been instantaneously affected. To do this, Darby et al. (3) leveraged a large reference dataset of restingstate functional MRI (fMRI) data from healthy individuals [from the Brain Genomics Superstruct Project (5)]. From each lesion patient a map was produced showing where, in this resting-state dataset of healthy individuals, the lesion site had strong functional connectivity. Maps of these functionally connected regions were then overlapped across the patients, producing a composite map showing all of the places in a healthy brain that were predicted to be impacted by the lesions. This innovative approach (4, 6) takes seriously the notion of a disconnection syndrome, a term popularized by Norman Geschwind (7) in the 1960s: it is the idea that behavioral impairments can be caused by the disruption of the connections between brain regions rather than by damage to one particular region itself.The composite connectivity map based on positive correlations in the resting-state dataset in fact included many of the regions that, when themselves lesioned, were associated with criminality: the ventromedial and orbi...

show abstract

“…persons, items, raters, time, or setting) and assess their relative 384 contribution. Future multi-site, multi-stimulus, multi-method, and population-level 385 initiatives may perform G analyses to produce comprehensive neural reliability maps for 386 different facets (Dubois and Adolphs, 2016;O'Connor et al, 2016). Related to this, the 387 time-series correlations we report are linked and in some cases mathematically 388 equivalent to intra-class correlations (ICC) (Shrout and Lane, 2012).…”

mentioning

confidence: 91%

“…399 patients), or different scanning parameters could change the magnitude of the 400 correlations. However, the outlined conceptual framework is robust to such specifics and 401 disentangling different facets of within and between person variation can contribute to 402 the development of fMRI in the next decades (Dubois and Adolphs, 2016).…”

mentioning

confidence: 99%

“…http://dx.doi.org/10.1101/158188 doi: bioRxiv preprint first posted online Jul. 3, 2017; important information about the distribution of normal brain function and for the 411 diagnosis of disorders (Dubois and Adolphs, 2016). Importantly, while this paper 412 focuses on methodological concepts, the measures of neural process similarity are of 413 interest in their own right and offer new means to examine inter-and intra-individual 414 neural differences in higher-order cognitive and motivational processes (Campbell et al, 415 2015;Hasson et al, 2009;Honey, Thompson, Lerner, and Hasson, 2012;Imhof et al, 416 2017; Naci, Cusack, Anello, and Owen, 2014; Schmälzle et al, 2013).…”

mentioning

confidence: 99%

See 1 more Smart Citation

Reliability of fMRI time series: Similarity of neural processing during movie viewing

Schmälzle

Imhof

Grall

et al. 2017

Preprint

View full text Add to dashboard Cite

Despite its widespread use in neuroscience, the reliability of fMRI remains insufficiently understood. One powerful way to tap into aspects of fMRI reliability is via the inter-subject correlation (ISC) approach, which exposes different viewers to the same time-locked naturalistic stimulus and assesses the similarity of neural time series. Here we examined the correlations of fMRI time series from 24 participants who watched the same movie clips across three repetitions. This enabled us to examine inter-subject correlations, intra-subject correlations, and correlations between aggregated time series, which we link to the notions of inter-rater reliability, stability, and consistency. In primary visual cortex we found average pairwise inter-subject correlations of about r = 0.3, and intra-subject correlations of similar magnitude. Aggregation across subjects increased inter-subject (inter-group) correlations to r = 0.87, and additional intra-subject averaging before cross-subject aggregation yielded correlations of r = 0.93. Computing the same analyses for parietal (visuospatial network) and cingulate cortices (saliency network) revealed a gradient of decreasing ISC from primary visual to higher visual to post-perceptual regions. These latter regions also benefitted most from the increased reliability due to aggregation. We discuss theoretical and practical implications of this link between neural process similarity and psychometric conceptions of inter-rater reliability, stability, and internal consistency.

show abstract

Building a Science of Individual Differences from fMRI

Cited by 570 publications

References 201 publications

Joint prediction of multiple scores captures better individual traits from brain images

Joint prediction of multiple scores captures better individual traits from brain images

Searching for the neural causes of criminal behavior

Reliability of fMRI time series: Similarity of neural processing during movie viewing

Contact Info

Product

Resources

About