Connectome-Based Predictive Modeling of Creativity Anxiety

Ren, Zhiting; Daker, Richard J.; Shi, Liang; Sun, Jiangzhou; Beaty, Roger E.; Wu, Xinran; Chen, Qunlin; Yang, Wenjing; Lyons, Ian M.; Green, Adam E.; Qiu, Jiang

doi:10.1016/j.neuroimage.2020.117469

Cited by 53 publications

(47 citation statements)

References 34 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Utilizing CPM, a number of studies have successfully predicted attention, anxiety, and mother-infant bonding (Ren et al, 2021;Rutherford et al, 2020;Yoo et al, 2018). In line with the conclusions of these prior studies, the present findings suggest that changes in functional connectivity may improve LS.…”

Section: Discussionsupporting

confidence: 86%

“…This research was inspired by connectome‐based predictive modeling (CPM), which uses large‐scale neuroimaging data to predict individual differences in traits and behavior (Shen et al., 2017 ). Utilizing CPM, a number of studies have successfully predicted attention, anxiety, and mother–infant bonding (Ren et al., 2021 ; Rutherford et al., 2020 ; Yoo et al., 2018 ). In line with the conclusions of these prior studies, the present findings suggest that changes in functional connectivity may improve LS.…”

Section: Discussionmentioning

confidence: 99%

See 1 more Smart Citation

Prediction of life satisfaction from resting‐state functional connectome

et al. 2021

View full text Add to dashboard Cite

Background: Better life satisfaction (LS) is associated with better psychological and psychiatric outcomes. To the best of our knowledge, no studies have examined prediction models for LS.Methods: Using resting-state functional magnetic resonance imaging (R-fMRI) data from the Human Connectome Project (HCP) Young Adult S1200 dataset, we examined whether LS is predictable from intrinsic functional connectivity (iFC). All the HCP data were subdivided into either discovery (n = 100) or validation (n = 766) datasets. Using R-fMRI data in the discovery dataset, we computed a matrix of iFCs between brain regions. Ridge regression, in combination with principal component analysis and 10fold cross-validation, was used to predict LS. Prediction performance was evaluated by comparing actual and predicted LS scores. The generalizability of the prediction model obtained from the discovery dataset was evaluated by applying this model to the validation dataset. Results:The model was able to successfully predict LS in the discovery dataset (r = 0.381, p < .001). The model was also able to successfully predict the degree of LS (r = 0.137, 5000-repetition permutation test p = .006) in the validation dataset, suggesting that our model is generalizable to the prediction of LS in young adults. iFCs stemming from visual, ventral attention, or limbic networks to other networks (such as the dorsal attention network and default mode network) were likely to contribute positively toward predicted LS scores. iFCs within ventral attention and limbic networks also positively contributed to predicting LS. On the other hand, iFCs stemming from the visual and cerebellar networks to other networks were likely to contribute negatively to the predicted LS scores. Conclusion:The present findings suggest that LS is predictable from the iFCs. These results are an important step toward identifying the neural basis of life satisfaction.

show abstract

Section: Discussionsupporting

confidence: 86%

Section: Discussionmentioning

confidence: 99%

Prediction of life satisfaction from resting‐state functional connectome

et al. 2021

View full text Add to dashboard Cite

show abstract

“…A lower MSE value means a small difference between predicted and observed score [30] . The signi cance of the constructed model was further tested by applying the 1000-permutation-test [15]. This test was done by randomly shu ing the UPDRS III score along with repeating the above processes 1000 times.…”

Section: Model Construction With Consensus Connections and Predictability Evaluationmentioning

confidence: 99%

“…Detecting the link between individual functional connectome and behavioral measurements can maximally reduce the bias brings from population variation, and the observed brain-behavior associations would increase robustness and generalizability [14]. Thus, a newly applied connectome-based predictive modeling (CPM) approach has been introduced to predict behavior at the individual level by using large-scale network functional connectivity in a machine-learning framework, which has been employed to explore complicated mechanisms in mental and cognitive disorders [15][16][17], as well as in predicting outcome after deep brain stimulation in PD patients [18].…”

mentioning

confidence: 99%

Identifying a Whole-brain Connectome-based Model in Drug-naïve Parkinson’s Disease for Predicting Motor Impairment

Zhou

Guo

et al. 2021

Preprint

View full text Add to dashboard Cite

Background The functional alternation of distinct brain networks contribute to motor impairment in Parkinson’s disease (PD) remains unclear. Identifying a whole-brain connectome-based predictive model (CPM) in drug-naïve patients and verifying its predictability among drug-managed patients would be helpful to detect generalizable brain-behavior association and reflect intrinsic functional underpinning of motor impairment. Methods Resting-state functional data of 47 drug-naïve patients were enrolled to construct a predictive model by using the CPM approach, which was subsequently validated in 115 drug-managed patients. The severity of motor impairment was assessed by calculating Unified Parkinson’s Disease Rating Scale part III (UPDRS III) scores. Predictive performance was evaluated with the correlation coefficient(rtrue) and the mean squared error (MSE) between observed and predicted scores. Results A CPM for predicting individual motor impairment in drug-naïve PD was identified with significant performance (rtrue=0.845, p < 0.001, MSE = 137.57). Two connection patterns were recognized according to the correlation coefficients between the connections’ strength and motor impairment severity. The negative motor-impairment-related network contained more within-network connections in the motor, visual-related, and default mode networks, while the positive motor-impairment-related network was constructed mostly with between-network connections coupled motor-visual, motor-limbic, and motor-basal ganglia networks. The predictability of constructed model was further confirmed in drug-managed patients (r = 0.209, p = 0.025, MSE = 182.96), suggesting generalizability in PD patients with lasting dopaminergic medication influence. Conclusions This study identified a whole-brain connectome-based model that could predict the severity of motor impairment for PD. The connection patterns generated from the model reflected that functional segregation of motor, visual-related, and default mode networks play an important role in PD motor impairment, and higher connections coupling motor and non-motor regions might demonstrate a compensatory mechanism to overcome motor impairment.

show abstract

“…Compared with conventional correlation or regression analyses, CPM sufficiently extracts brain connectivity data to build predictive models and employs cross-validation to protect against overfitting (Shen et al, 2017; Yip, Scheinost, Potenza, & Carroll, 2019). This method has been well-validated in generating predictive model based on neural features (Feng, Wang, Li, & Xu, 2019; Ren et al, 2020). For example, CPM has been used to predict temperament (Jiang et al, 2018) and propensity to trust (Lu et al, 2019) based on whole-brain RSFC.…”

Section: Introductionmentioning

confidence: 99%

Connectome-based model predicts individual psychopathic traits in college students

Zhu²,

Zhao

et al. 2021

Preprint

View full text Add to dashboard Cite

Background: Psychopathic traits have been suggested to increase the risk of violations of socio-moral norms. Previous studies revealed that abnormal neural signatures are associated with elevated psychopathic traits; however, whether the intrinsic network architecture can predict psychopathic traits at the individual level remains unclear. Methods: The present study utilized connectome-based predictive modeling (CPM) to investigate whether whole-brain resting-state functional connectivity (RSFC) can predict psychopathic traits in the general population. RS functional magnetic resonance imaging data were collected from 84 college students with varying psychopathic traits measured by the Levenson Self-Report Psychopathy Scale (LSRP). Results: We found that RSFC of the negative networks predicted individual differences in total LSRP and secondary psychopathy scores but not primary psychopathy score. Particularly, nodes with the most connections in the predictive connectome anchored in the prefrontal cortex (e.g., anterior prefrontal cortex and orbitofrontal cortex) and limbic system (e.g., anterior cingulate cortex and insula). In addition, the connections between the occipital network (OCCN) and cingulo-opercular network (CON) served as a significant predictive connectome for total LSRP and secondary psychopathy score. Conclusion: CPM constituted by whole-brain RSFC significantly predicted psychopathic traits individually in the general population. The prefrontal cortex and limbic system at the anatomic level and the CON and OCCN at the functional network level plays a special role in the predictive model-reflecting atypical executive control and affective processing for individuals with elevated psychopathic traits. These findings may provide some implications for early detection and potential intervention of psychopathic tendency.

show abstract

Connectome-Based Predictive Modeling of Creativity Anxiety

Cited by 53 publications

References 34 publications

Prediction of life satisfaction from resting‐state functional connectome

Prediction of life satisfaction from resting‐state functional connectome

Identifying a Whole-brain Connectome-based Model in Drug-naïve Parkinson’s Disease for Predicting Motor Impairment

Connectome-based model predicts individual psychopathic traits in college students

Contact Info

Product

Resources

About