Using a Machine Learning Mega-Analysis to Predict Alcohol Use Disorder
Poster No:
538
Submission Type:
Abstract Submission
Authors:
McKinney Pitts1, Anthony Juliano1, Tony Barrows1, Nicholas Allgaier1, Dekang Yuan1, Hugh Garavan1
Institutions:
1University of Vermont College of Medicine, Burlington, VT
First Author:
Co-Author(s):
Introduction:
Alcohol use disorder (AUD) is a medical condition that heavily impacts the lives of those who struggle with it as well as the lives of those around them. It is frequently characterized by profound detrimental effects on social, financial, and physical wellbeing. Understanding the neural basis of AUD has the potential to inform our understanding of this pervasive disorder and develop better models of treatment. Recent works have implemented a variety of methods including machine learning approaches to investigate risk factors, remission, and comorbidities relating to AUD to further our understanding of the neurobiology of this condition (Kinreich et al., 2021; Lee et al., 2019; Mackey et al., 2019; see Ebrahimi et al., 2021 for review). The current study adds to this literature by investigating the neural correlates of AUD with a notably large sample size and subsequent statistical power.
Methods:
In this study, we conducted a mega-analysis of datasets pooled from 28 studies collected by the ENIGMA Addiction Working Group (n = 2,253, 1,947 males, 1,781 cases). We utilized elastic net regression to distinguish cases from controls using measures of cortical thickness and surface area and subcortical volumes in a single regression model. Regions of interest were defined from the Desikan-Killany atlas (Desikan et al., 2006), and scanner effects were removed from the data using ComBat (Johnson et al., 2007; Yu et al., 2018) with age, sex, and intracranial volume included as covariates. To assess the generalizability of the multivariate model, a 5-fold cross-validation scheme with nested hyperparameter tuning was implemented.
Results:
The final model performed with greater than 70% accuracy and revealed several key systems involved in distinguishing AUD such as: frontal regions (cortical thickness: left lateral orbitofrontal, right frontal pole, left superior frontal, right pars opercularis, and right pars orbitalis; surface area: right caudal middle frontal and right rostral middle frontal), parietal and occipital cortex (cortical thickness: left cuneus, left inferior parietal, bilateral lingual, and bilateral left superior parietal; surface area: left precuneus), cingulate cortex (cortical thickness: right caudal anterior cingulate, right posterior cingulate, and right rostral anterior cingulate), subcortical areas (volume: bilateral amygdala, left caudate, and right putamen).
·ROC curve of the final model
Conclusions:
The results reveal that a broad range of brain regions contribute to the successful brain-based identification of AUD. The findings show substantial overlap between the regions driving the successful classification and prior results of alcohol-related changes in brain structure. Revealing these regions has potential utility for assessing structural markers of recovery thereby informing the efficacy of treatment plans. Additionally, a multivariate classifier of AUD can assess risk for AUD if coupled with increased polygenic risk scores for alcoholism or familial history. This data may also provide information to help identify subtypes of AUD as seen in previous work (Cao et al., 2022). By utilizing the large amount of data in the present study, more nuanced patterns of brain changes may suggest further classification into AUD subtypes. This approach also enables us to investigate the brain correlates of other substance use disorders and mental health conditions.
Disorders of the Nervous System:
Psychiatric (eg. Depression, Anxiety, Schizophrenia) 1
Modeling and Analysis Methods:
Classification and Predictive Modeling 2
Neuroanatomy, Physiology, Metabolism and Neurotransmission:
Cortical Anatomy and Brain Mapping
Subcortical Structures
Neuroinformatics and Data Sharing:
Databasing and Data Sharing
Keywords:
Addictions
Cortex
Data analysis
Machine Learning
MRI
Sub-Cortical
1|2Indicates the priority used for review
Abstract Information
By submitting your proposal, you grant permission for the Organization for Human Brain Mapping (OHBM) to distribute your work in any format, including video, audio print and electronic text through OHBM OnDemand, social media channels, the OHBM website, or other electronic publications and media.
The Open Science Special Interest Group (OSSIG) is introducing a reproducibility challenge for OHBM 2025. This new initiative aims to enhance the reproducibility of scientific results and foster collaborations between labs. Teams will consist of a “source” party and a “reproducing” party, and will be evaluated on the success of their replication, the openness of the source work, and additional deliverables. Click here for more information. Propose your OHBM abstract(s) as source work for future OHBM meetings by selecting one of the following options:
Please indicate below if your study was a "resting state" or "task-activation” study.
Healthy subjects only or patients (note that patient studies may also involve healthy subjects):
Was this research conducted in the United States?
Are you Internal Review Board (IRB) certified? Please note: Failure to have IRB, if applicable will lead to automatic rejection of abstract.
Were any human subjects research approved by the relevant Institutional Review Board or ethics panel? NOTE: Any human subjects studies without IRB approval will be automatically rejected.
Were any animal research approved by the relevant IACUC or other animal research panel? NOTE: Any animal studies without IACUC approval will be automatically rejected.
Please indicate which methods were used in your research:
For human MRI, what field strength scanner do you use?
Which processing packages did you use for your study?
Provide references using APA citation style.
2. Desikan, R. S., Ségonne, F., Fischl, B., Quinn, B. T., Dickerson, B. C., Blacker, D., ... & Killiany, R. J. (2006). An automated labeling system for subdividing the human cerebral cortex on MRI scans into gyral based regions of interest. Neuroimage, 31(3), 968-980.
3. Ebrahimi, A., Wiil, U. K., Schmidt, T., Naemi, A., Nielsen, A. S., Shaikh, G. M., & Mansourvar, M. (2021). Predicting the risk of alcohol use disorder using machine learning: a systematic literature review. IEEE Access, 9, 151697-151712.
4. Johnson, W. E., Li, C., & Rabinovic, A. (2007). Adjusting batch effects in microarray expression data using empirical Bayes methods. Biostatistics, 8(1), 118-127.
5. Kinreich, S., McCutcheon, V. V., Aliev, F., Meyers, J. L., Kamarajan, C., Pandey, A. K., ... & Porjesz, B. (2021). Predicting alcohol use disorder remission: a longitudinal multimodal multi-featured machine learning approach. Translational psychiatry, 11(1), 166.
6. Lee, M. R., Sankar, V., Hammer, A., Kennedy, W. G., Barb, J. J., McQueen, P. G., & Leggio, L. (2019). Using machine learning to classify individuals with alcohol use disorder based on treatment seeking status. EClinicalMedicine, 12, 70-78.
7. Mackey, S., Allgaier, N., Chaarani, B., Spechler, P., Orr, C., Bunn, J., ... & ENIGMA Addiction Working Group. (2019). Mega-analysis of gray matter volume in substance dependence: general and substance-specific regional effects. American Journal of Psychiatry, 176(2), 119-128.
8. Yu, M., Linn, K. A., Cook, P. A., Phillips, M. L., McInnis, M., Fava, M., ... & Sheline, Y. I. (2018). Statistical harmonization corrects site effects in functional connectivity measurements from multi‐site fMRI data. Human brain mapping, 39(11), 4213-4227.