Format

Send to

Choose Destination
Biol Psychiatry Cogn Neurosci Neuroimaging. 2019 Jun;4(6):554-566. doi: 10.1016/j.bpsc.2019.04.013. Epub 2019 May 10.

Functional and Optogenetic Approaches to Discovering Stable Subtype-Specific Circuit Mechanisms in Depression.

Author information

1
Feil Family Brain and Mind Research Institute and Department of Psychiatry, Weill Cornell Medicine, New York, New York; Department of Statistics, Columbia University, New York, New York; Simons Foundation, New York, New York.
2
Feil Family Brain and Mind Research Institute and Department of Psychiatry, Weill Cornell Medicine, New York, New York.
3
Department of Psychiatry, Toronto Western Hospital, Toronto, Ontario, Canada.
4
Feil Family Brain and Mind Research Institute and Department of Psychiatry, Weill Cornell Medicine, New York, New York. Electronic address: col2004@med.cornell.edu.

Abstract

BACKGROUND:

Previously, we identified four depression subtypes defined by distinct functional connectivity alterations in depression-related brain networks, which in turn predicted clinical symptoms and treatment response. Optogenetic functional magnetic resonance imaging offers a promising approach for testing how dysfunction in specific circuits gives rise to subtype-specific, depression-related behaviors. However, this approach assumes that there are robust, reproducible correlations between functional connectivity and depressive symptoms-an assumption that was not extensively tested in previous work.

METHODS:

First, we comprehensively reevaluated the stability of canonical correlations between functional connectivity and symptoms (N = 220 subjects) using optimized approaches for large-scale statistical hypothesis testing, and we validated methods for improving estimation of latent variables driving brain-behavior correlations. Having confirmed this necessary condition, we reviewed recent advances in optogenetic functional magnetic resonance imaging and illustrated one approach to formulating hypotheses regarding latent subtype-specific circuit mechanisms and testing them in animal models.

RESULTS:

Correlations between connectivity features and clinical symptoms were robustly significant, and canonical correlation analysis solutions tested repeatedly on held-out data generalized. However, they were sensitive to data quality, preprocessing, and clinical heterogeneity, which can reduce effect sizes. Generalization could be markedly improved by adding L2 regularization, which decreased estimator variance, increased canonical correlations in left-out data, and stabilized feature selection. These improvements were useful for identifying candidate circuits for optogenetic interrogation in animal models.

CONCLUSIONS:

Multiview, latent-variable approaches such as canonical correlation analysis offer a conceptually useful framework for discovering stable patient subtypes by synthesizing multiple clinical and functional measures. Optogenetic functional magnetic resonance imaging holds promise for testing hypotheses regarding latent, subtype-specific mechanisms driving depressive symptoms and behaviors.

KEYWORDS:

Biomarkers; Depression; Depression subtypes; Machine learning; Neuroimaging; Optogenetic fMRI

PMID:
31176387
PMCID:
PMC6788795
[Available on 2020-06-01]
DOI:
10.1016/j.bpsc.2019.04.013

Supplemental Content

Full text links

Icon for Elsevier Science
Loading ...
Support Center