[WIP] Review: connectome-based predictive model
Last updated: April 190, 19049
Review of connectome-based predictive modeling
In the search of pubmed, couple of papers were found. In this report, we review some of the recent works in prediction models using connectome derived from non-trask fMRI data.
Prediction performance
1. Dadi et al. (2019) paperlink
- evaluate methods for prediction from resting-state fMRI on 6 cohorts.
- A prediction pipeline needs brain regions, a connectome, and a supervised predictor.
- Regions defined functionally (with dictionary learning or ICA) give best prediction.
- Prefer tangent-space parametrization of connectomes to full or partial correlation.
- Non-sparse linear classifiers are best for supervised learning.
2. Jiang et al. (2018) paperlink
- Development of connectome-based predictive models for four temperament metrics: p-values based thresholding on the correlation matrix, and LASSO.
- Temperaments can be characterized by functional connections within the frontal-subcortical circuits.
- Prediction of harm avoidance can be generalized to neuroticism in new independent dataset.
- Prediction of novelty seeking can be generalized to extraversion in new independent dataset.
3. Lu et al. (2019) paperlink
- Power’s atlas (in MNI space) with kernel regression.
4. Fountain-Zaragoza et al. (2019) paperlink
- 268-node Shen functional atlas
- Use same prediction method as Rosenberg et al. (2018) Seems to take average of thresholded connections (need to check..I generally do not agree with this..)
- This paper did not do actual prediction. Their small sample size is inadequate to evaluate prediction (n=34), anyways.
5. Rosenberg et al. (2018) paperlink
CPM
- Briefly, network nodes were defined with the Shen 268-node functional brain atlas
- Conduct leave-one-out cross-validation.
- In each inner loop, spearman’s correlation between behavioral outcomes and each of the connecitivities (z-scores of the correlation) with 5% threshold and takes the average of the z-scores by positive and negative correlation, separately.
- Linear regression will be used for predition model.
- Taking average….well…we should go by network otherwise it’s hard to figure out.. potentially oversimplification.
Other papers to review
(Yoo et al. 2018) (Feng et al. 2018) (Lin et al. 2018) (Beaty et al. 2018) (Shen et al. 2017)
References
Beaty, Roger E, Yoed N Kenett, Alexander P Christensen, Monica D Rosenberg, Mathias Benedek, Qunlin Chen, Andreas Fink, et al. 2018. “Robust Prediction of Individual Creative Ability from Brain Functional Connectivity.” Proceedings of the National Academy of Sciences 115 (5):1087–92.
Dadi, Kamalaker, Mehdi Rahim, Alexandre Abraham, Darya Chyzhyk, Michael Milham, Bertrand Thirion, Gaël Varoquaux, Alzheimer’s Disease Neuroimaging Initiative, and others. 2019. “Benchmarking Functional Connectome-Based Predictive Models for Resting-State fMRI.” NeuroImage.
Feng, Chunliang, Jie Yuan, Haiyang Geng, Ruolei Gu, Hui Zhou, Xia Wu, and Yuejia Luo. 2018. “Individualized Prediction of Trait Narcissism from Whole-Brain Resting-State Functional Connectivity.” Human Brain Mapping 39 (9):3701–12.
Fountain-Zaragoza, Stephanie, Shaadee Samimy, Monica D Rosenberg, and Ruchika Shaurya Prakash. 2019. “Connectome-Based Models Predict Attentional Control in Aging Adults.” NeuroImage 186:1–13.
Jiang, Rongtao, Vince D Calhoun, Nianming Zuo, Dongdong Lin, Jin Li, Lingzhong Fan, Shile Qi, et al. 2018. “Connectome-Based Individualized Prediction of Temperament Trait Scores.” NeuroImage 183:366–74.
Lin, Qi, Monica D Rosenberg, Kwangsun Yoo, Tiffany W Hsu, Thomas P O’Connell, and Marvin M Chun. 2018. “Resting-State Functional Connectivity Predicts Cognitive Impairment Related to Alzheimer’s Disease.” Frontiers in Aging Neuroscience 10:94.
Lu, Xiaping, Ting Li, Zhichao Xia, Ruida Zhu, Li Wang, Yue-Jia Luo, Chunliang Feng, and Frank Krueger. 2019. “Connectome-Based Model Predicts Individual Differences in Propensity to Trust.” Human Brain Mapping 40 (6):1942–54.
Rosenberg, Monica D, Wei-Ting Hsu, Dustin Scheinost, R Todd Constable, and Marvin M Chun. 2018. “Connectome-Based Models Predict Separable Components of Attention in Novel Individuals.” Journal of Cognitive Neuroscience 30 (2):160–73.
Shen, Xilin, Emily S Finn, Dustin Scheinost, Monica D Rosenberg, Marvin M Chun, Xenophon Papademetris, and R Todd Constable. 2017. “Using Connectome-Based Predictive Modeling to Predict Individual Behavior from Brain Connectivity.” Nature Protocols 12 (3):506.
Yoo, Kwangsun, Monica D Rosenberg, Wei-Ting Hsu, Sheng Zhang, Chiang-Shan R Li, Dustin Scheinost, R Todd Constable, and Marvin M Chun. 2018. “Connectome-Based Predictive Modeling of Attention: Comparing Different Functional Connectivity Features and Prediction Methods Across Datasets.” Neuroimage 167:11–22.