Skip to main page content
U.S. flag

An official website of the United States government

Dot gov

The .gov means it’s official.
Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.

Https

The site is secure.
The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
. 2023 Mar;8(3):238-240.
doi: 10.1016/j.bpsc.2023.01.003.

Deriving Generalizable and Interpretable Brain-Behavior Phenotypes of Cannabis Use

Anna B Konova  1 , Anna Zilverstand  2
Affiliations

Deriving Generalizable and Interpretable Brain-Behavior Phenotypes of Cannabis Use

Anna B Konova et al. Biol Psychiatry Cogn Neurosci Neuroimaging. 2023 Mar.
No abstract available

PubMed Disclaimer

Conflict of interest statement

The authors report no biomedical financial interests or potential conflicts of interest.

Figures

Figure 1.
Figure 1.
Idealized analytic pipeline for deriving generalizable and interpretable brain-behavior phenotypes of cannabis use (and other psychiatric disorders). To meet the sample size demands of modern machine learning approaches, data pooling of deeply phenotyped and diverse samples is encouraged. After determining the clinical outcome of interest, initial steps should focus on 1) determining the relevant data input based on a combination of theory and quantitative tests (e.g., task vs. rest, activity vs. connectivity), 2) determining the most appropriate validation scheme for the data size and type, and 3) identifying the best model (via formal model selection and hyperparameter optimization) for the chosen input data. Once the best model is identified based on appropriate performance metrics and out-of-sample accuracy is determined, 4) a series of steps should be taken to interrogate and interpret the model. This may include the mapping of top-ranked features that contribute to model performance (e.g., ranking by bootstrapping, weights, or complex tools such as Shapley additive explanations), as well as determining the influence of sample characteristics including based on variables deemed to be nuisance variables in addition to variables of interest. [Figure conceptually adapted from (2,5).] Icons made by Vitaly Gorbachev, Freepik, and Justicon from www.flaticon.com.
See this image and copyright information in PMC

Comment on

  • Biol Psychiatry Cogn Neurosci Neuroimaging.

References

    1. Connor JP, Stjepanovic D, Le Foll B, Hoch E, Budney AJ, Hall WD (2021): Cannabis use and cannabis use disorder. Nat Rev Dis Primers 7:16. - PMC - PubMed
    1. Kulkarni KR, Schafer M, Berner LA, Fiore VG, Heflin M, Hutchison K, et al. (2023): An interpretable and predictive connectivity-based neural signature for chronic cannabis use. Biol Psychiatry Cogn Neurosci Neuroimaging 8:320–330. - PMC - PubMed
    1. Dhamala E, Yeo BTT, Holmes AJ (2022): One size does not fit all: Methodological considerations for brain-based predictive modeling in psychiatry [published online Sep 29]. Biol Psychiatry. - PubMed
    1. Kohoutová L, Heo J, Cha S, Lee S, Moon T, Wager TD, Woo C-W (2020): Toward a unified framework for interpreting machine-learning models in neuroimaging. Nat Protoc 15:1399–1435. - PMC - PubMed
    1. Scheinost D, Noble S, Horien C, Greene AS, Lake EM, Salehi M, et al. (2019): Ten simple rules for predictive modeling of individual differences in neuroimaging. Neuroimage 193:35–45. - PMC - PubMed

Publication types