Connectome-based predictive modeling of cognitive reserve using task-based functional connectivity

Boyle, Rory; Connaughton, Michael; McGlinchey, Eimear; Knight, Silvin P.; Looze, Céline De; Carey, Daniel; Stern, Yaakov; Robertson, Ian H.; Kenny, Rose Anne; Whelan, Robert

doi:10.1101/2022.06.01.494342

Cited by 3 publications

(3 citation statements)

References 92 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Cognitive deficits are associated with a range of psychiatric and developmental disorders (Millan et al, 2012;Zelazo, 2020). Achieving robust predictions of these constructs is meaningful for cognitive and clinical neuroscience (Barron et al, 2020;Boyle et al, 2023;Sui et al, 2020). However, the observed effect sizes are still smaller than necessary for real-world utility.…”

Section: Discussionmentioning

confidence: 99%

Brain-phenotype predictions can survive across diverse real-world data

Adkinson,

Rosenblatt,

Dadashkarimi

et al. 2024

Preprint

View full text Add to dashboard Cite

Recent work suggests that machine learning models predicting psychiatric treatment outcomes based on clinical data may fail when applied to unharmonized samples. Neuroimaging predictive models offer the opportunity to incorporate neurobiological information, which may be more robust to dataset shifts. Yet, among the minority of neuroimaging studies that undertake any form of external validation, there is a notable lack of attention to generalization across dataset-specific idiosyncrasies. Research settings, by design, remove the between-site variations that real-world and, eventually, clinical applications demand. Here, we rigorously test the ability of a range of predictive models to generalize across three diverse, unharmonized samples: the Philadelphia Neurodevelopmental Cohort (n=1291), the Healthy Brain Network (n=1110), and the Human Connectome Project in Development (n=428). These datasets have high inter-dataset heterogeneity, encompassing substantial variations in age distribution, sex, racial and ethnic minority representation, recruitment geography, clinical symptom burdens, fMRI tasks, sequences, and behavioral measures. We demonstrate that reproducible and generalizable brain-behavior associations can be realized across diverse dataset features with sample sizes in the hundreds. Results indicate the potential of functional connectivity-based predictive models to be robust despite substantial inter-dataset variability. Notably, for the HCPD and HBN datasets, the best predictions were not from training and testing in the same dataset (i.e., cross-validation) but across datasets. This result suggests that training on diverse data may improve prediction in specific cases. Overall, this work provides a critical foundation for future work evaluating the generalizability of neuroimaging predictive models in real-world scenarios and clinical settings.

show abstract

Section: Discussionmentioning

confidence: 99%

Brain-phenotype predictions can survive across diverse real-world data

Adkinson,

Rosenblatt,

Dadashkarimi

et al. 2024

Preprint

View full text Add to dashboard Cite

show abstract

“…We aim to 1) determine if attention-based predictive models can be generated in a neurodiverse developmental dataset, 2) test if such a model generalizes out of sample, and 3) interrogate neuroanatomy of the network model and assess the stability in individual participants across time. Using connectome-based predictive modelling (CPM) (Finn et al 2015; Rosenberg, Finn, et al 2016; Shen et al 2017; Beaty et al 2018; Greene et al 2018; Hsu et al 2018; Yoo et al 2018; Rapuano et al 2020; Rohr et al 2020; Boyle et al 2022), we show that we are indeed able to predict performance on an in-scan sustained attention task in novel subjects based on functional connectivity data. The predictions are robust to factors such as in-scanner head motion, Autism Diagnostic Observation Schedule (ADOS) scores, age, sex, and intelligence quotient (IQ) scores.…”

Section: Introductionmentioning

confidence: 96%

A generalizable connectome-based marker of in-scan sustained attention in neurodiverse youth

Horien

Greene

Shen

et al. 2022

Preprint

View full text Add to dashboard Cite

Difficulty with attention is an important symptom in many conditions in psychiatry, including neurodiverse conditions such as autism. There is a need to better understand the neurobiological correlates of attention and leverage these findings for individuals in healthcare settings. Nevertheless, it remains unclear if it is possible to build robust dimensional predictive models of attention in neurodiverse populations. Here, we use five datasets to identify and validate functional connectome-based markers of attention. In dataset one, we use connectome-based predictive modelling and observe successful prediction of performance on an in-scan sustained attention task in a neurodiverse sample of youth. The predictions are not driven by confounds, such as head motion. In dataset two, we find the attention network model defined in dataset one generalizes to predict in-scan attention in a separate sample of neurotypical participants performing the same attention task. In datasets three to five, we use connectome-based identification and longitudinal scans to probe the stability of the attention network across months to years in individual participants. Our results help elucidate the brain correlates of attention in neurodiverse youth and support the further development of predictive dimensional models of other clinically-relevant phenotypes.

show abstract

“…CPM has been applied to successfully predict individual differences in various cognitive phenotypes including global cognition (Lin et al, 2018), attention (Fountain-Zaragoza et al, 2019;Rosenberg et al, 2016), executive function (Henneghan et al, 2020), fluid intelligence (Finn et al, 2015;S. Gao et al, 2019;Greene et al, 2018), processing speed (M. Gao et al, 2020), cognitive reserve (Boyle et al, 2022), and creative ability (Beaty et al, 2018). CPM has also enabled accurate prediction of individual differences in behavioural phenotypes such as anxiety (Wang et al, 2021), depression (Ju et al, 2020), feelings of loneliness (Feng et al, 2019) and stress (Goldfarb et al, 2020), childhood aggression (Ibrahim et al, 2021) and social impairments (Dufford et al, 2022), and abstinence from use of substances including opioids (Lichenstein et al, 2019) and cocaine (Yip et al, 2019).…”

Section: Introductionmentioning

confidence: 99%

Studying the connectome at a large scale

Boyle¹,

Weng²

2023

Preprint

View full text Add to dashboard Cite

This chapter outlines a flexible connectome-based predictive modeling method that is optimised for large neuroimaging datasets via the use of parallel computing and by adding the capability to account for possible site- and scanner-related heterogeneity in multi-site neuroimaging datasets. We present the decision points that need to be made when conducting a connectome-based predictive modeling analysis and we provide full code to conduct an analysis on public data. To date, connectome-based predictive modeling has been applied to predict different cognitive and behavioural phenotypes with many studies reporting accurate predictions that generalized to external datasets.

show abstract

Connectome-based predictive modeling of cognitive reserve using task-based functional connectivity

Cited by 3 publications

References 92 publications

Brain-phenotype predictions can survive across diverse real-world data

Brain-phenotype predictions can survive across diverse real-world data

A generalizable connectome-based marker of in-scan sustained attention in neurodiverse youth

Studying the connectome at a large scale

Contact Info

Product

Resources

About