Protocol | Published:

Using connectome-based predictive modeling to predict individual behavior from brain connectivity

Nature Protocols volume 12, pages 506518 (2017) | Download Citation


Neuroimaging is a fast-developing research area in which anatomical and functional images of human brains are collected using techniques such as functional magnetic resonance imaging (fMRI), diffusion tensor imaging (DTI), and electroencephalography (EEG). Technical advances and large-scale data sets have allowed for the development of models capable of predicting individual differences in traits and behavior using brain connectivity measures derived from neuroimaging data. Here, we present connectome-based predictive modeling (CPM), a data-driven protocol for developing predictive models of brain–behavior relationships from connectivity data using cross-validation. This protocol includes the following steps: (i) feature selection, (ii) feature summarization, (iii) model building, and (iv) assessment of prediction significance. We also include suggestions for visualizing the most predictive features (i.e., brain connections). The final result should be a generalizable model that takes brain connectivity data as input and generates predictions of behavioral measures in novel subjects, accounting for a considerable amount of the variance in these measures. It has been demonstrated that the CPM protocol performs as well as or better than many of the existing approaches in brain–behavior prediction. As CPM focuses on linear modeling and a purely data-driven approach, neuroscientists with limited or no experience in machine learning or optimization will find it easy to implement these protocols. Depending on the volume of data to be processed, the protocol can take 10–100 min for model building, 1–48 h for permutation testing, and 10–20 min for visualization of results.

Access optionsAccess options

Rent or Buy article

Get time limited or full article access on ReadCube.


All prices are NET prices.


  1. 1.

    , , & Circular analysis in systems neuroscience: the dangers of double dipping. Nat. Neurosci. 12, 535–540 (2009).

  2. 2.

    , , & Puzzlingly high correlations in fMRI studies of emotion, personality, and social cognition. Perspect. Psychol. Sci. 4, 274–290 (2009).

  3. 3.

    , & Prediction as a humanitarian and pragmatic contribution from human cognitive neuroscience. Neuron 85, 11–26 (2015).

  4. 4.

    et al. Functional connectome fingerprinting: identifying individuals using patterns of brain connectivity. Nat. Neurosci. 18, 1664–1671 (2015).

  5. 5.

    et al. A neuromarker of sustained attention from whole-brain functional connectivity. Nat. Neurosci. 19, 165–171 (2016).

  6. 6.

    et al. The WU-Minn human connectome project: an overview. Neuroimage 80, 62–79 (2013).

  7. 7.

    et al. The NKI-Rockland Sample: a model for accelerating the pace of discovery science in psychiatry. Front. Neurosci. 6, 152 (2012).

  8. 8.

    , , & The ADHD-200 consortium: a model to advance the translational potential of neuroimaging in clinical neuroscience. Front. Syst. Neurosci. 6, 62 (2012).

  9. 9.

    et al. The Philadelphia Neurodevelopmental Cohort: a publicly available resource for the study of normal and abnormal brain development in youth. Neuroimage 124, 1115–1119 (2016).

  10. 10.

    , , , & Parcellations and hemispheric asymmetries of human cerebral cortex analyzed on surface-based atlases. Cereb. Cortex 22, 2241–2262 (2012).

  11. 11.

    , , & Groupwise whole-brain parcellation from resting-state fMRI data for network node identification. Neuroimage 82, 403–415 (2013).

  12. 12.

    et al. Functional network organization of the human brain. Neuron 72, 665–678 (2011).

  13. 13.

    , , , & A whole brain fMRI atlas generated via spatially constrained spectral clustering. Hum. Brain Mapp. 33, 1914–1928 (2012).

  14. 14.

    & When optimism hurts: inflated predictions in psychiatric neuroimaging. Biol. Psychiatry 75, 746–748 (2014).

  15. 15.

    et al. HCP beta-release of the Functional Connectivity MegaTrawl. (2015).

  16. 16.

    et al. Functional connectomics from resting-state fMRI. Trends Cogn. Sci. 17, 666–682 (2013).

  17. 17.

    & Support vector regression machines. Adv. Neural Inf. Process. Syst. 9, 155–161 (1997).

  18. 18.

    et al. Prediction of individual brain maturity using fMRI. Science 329, 1358–1361 (2010).

  19. 19.

    et al. ADHD-200 global competition: diagnosing ADHD using personal characteristic data can outperform resting state fMRI measurements. Front. Syst. Neurosci. 6, 69 (2012).

  20. 20.

    , & Functional connectivity classification of autism identifies highly predictive brain features but falls short of biomarker standards. NeuroImage Clin. 7, 359–366 (2015).

  21. 21.

    et al. Functional connectivity magnetic resonance imaging classification of autism. Brain 134, 3742–3754 (2011).

  22. 22.

    , , & Classification of schizophrenia patients based on resting-state functional network connectivity. Front. Neurosci. 7, 133 (2013).

  23. 23.

    , & Identifying patients with Alzheimer's disease using resting-state fMRI and graph theory. Clin. Neurophysiol. 126, 2132–2141 (2015).

  24. 24.

    et al. Identifying major depression using whole-brain functional connectivity: a multivariate pattern analysis. Brain 135, 1498–1507 (2012).

  25. 25.

    et al. On the interpretation of weight vectors of linear models in multivariate neuroimaging. Neuroimage 87, 96–110 (2014).

  26. 26.

    , & The influence of head motion on intrinsic functional connectivity MRI. Neuroimage 59, 431–438 (2012).

  27. 27.

    , , , & Spurious group differences due to head motion in a diffusion MRI study. Neuroimage 88C, 79–90 (2013).

  28. 28.

    , & Recent progress and outstanding issues in motion correction in resting state fMRI. Neuroimage 105, 536–551 (2015).

  29. 29.

    et al. A comprehensive assessment of regional variation in the impact of head micromovements on functional connectomics. Neuroimage 76, 183–201 (2013).

  30. 30.

    . A study of cross-validation and bootstrap for accuracy estimation and model selection. in Proceedings of the 14th International Joint Conference on Artificial Intelligence 2, 1137–1143 (1995).

  31. 31.

    , & Assessing the generalizability of prognostic information. Ann. Intern. Med. 130, 515–524 (1999).

  32. 32.

    et al. Assessing the performance of prediction models: a framework for traditional and novel measures. Epidemiology 21, 128–138 (2010).

  33. 33.

    & Nonparametric Statistical Inference (Springer, 2011).

  34. 34.

    & Robust regression using iteratively reweighted least-squares. Commun. Stat.-Theory Methods 6, 813–827 (1977).

  35. 35.

    , & A note on computing robust regression estimates via iteratively reweighted least squares. Am. Stat. 42, 152–154 (1988).

  36. 36.

    et al. The organization of the human cerebral cortex estimated by intrinsic functional connectivity. J. Neurophysiol. 106, 1125–1165 (2011).

  37. 37.

    et al. Resting-state functional connectivity predicts impulsivity in economic decision-making. J. Neurosci. 33, 4886–4895 (2013).

  38. 38.

    et al. Neural predictors of individual differences in response to math tutoring in primary-grade school children. Proc. Natl. Acad. Sci. USA 110, 8230–8235 (2013).

  39. 39.

    et al. Predicting treatment response in social anxiety disorder from functional magnetic resonance imaging. JAMA Psychiatry 70, 87–97 (2013).

  40. 40.

    , & Structural maturation and brain activity predict future working memory capacity during childhood development. J. Neurosci. 34, 1592–1598 (2014).

Download references


M.D.R. and E.S.F. are supported by US National Science Foundation Graduate Research Fellowships. This work was also supported by US National Institutes of Health grant EB009666 to R.T.C.

Author information


  1. Department of Radiology and Biomedical Imaging, Yale School of Medicine, New Haven, Connecticut, USA.

    • Xilin Shen
    • , Dustin Scheinost
    • , Xenophon Papademetris
    •  & R Todd Constable
  2. Interdepartmental Neuroscience Program, Yale School of Medicine, New Haven, Connecticut, USA.

    • Emily S Finn
    • , Marvin M Chun
    •  & R Todd Constable
  3. Department of Psychology, Yale University, New Haven, Connecticut, USA.

    • Monica D Rosenberg
    •  & Marvin M Chun
  4. Department of Neuroscience, Yale School of Medicine, New Haven, Connecticut, USA.

    • Marvin M Chun
  5. Department of Biomedical Engineering, Yale University, New Haven, Connecticut, USA.

    • Xenophon Papademetris
  6. Department of Neurosurgery, Yale School of Medicine, New Haven, Connecticut, USA.

    • R Todd Constable


  1. Search for Xilin Shen in:

  2. Search for Emily S Finn in:

  3. Search for Dustin Scheinost in:

  4. Search for Monica D Rosenberg in:

  5. Search for Marvin M Chun in:

  6. Search for Xenophon Papademetris in:

  7. Search for R Todd Constable in:


X.S., E.S.F., D.S., X.P., and R.T.C. conceptualized the study. X.S. developed this protocol with help from E.S.F. and D.S. E.S.F. developed the prediction framework with help from X.S. and M.D.R. E.S.F., X.P., and X.S. contributed previously unpublished tools. X.P. developed the online visualization tools with help from X.S. and D.S. X.P., M.M.C., and R.T.C. provided support and guidance with data interpretation. All authors made valuable comments on the manuscript.

Competing interests

X.P. is a consultant for Electrical Geodesics Inc.

Corresponding author

Correspondence to R Todd Constable.

Supplementary information

PDF files

  1. 1.

    Supplementary Note

    Performance comparison between CPM- and SVR-based methods.

About this article

Publication history



Further reading


By submitting a comment you agree to abide by our Terms and Community Guidelines. If you find something abusive or that does not comply with our terms or guidelines please flag it as inappropriate.