EUVIMEDEUVIMED entdecken
Open Access

Classification of Alcohol Use Disorder With Explainable AI

Alexander S Hatoum iD, Alex P Miller iD, Hunter L Mathews iD, Aaron J Gorelik iD, David A A Baranger, Zhen Luo, Jodi Kutzer, Emma C Johnson iD, Arpana Agrawal, Ryan Bogdan

Alcohol, Clinical and Experimental Research · 2026

DOI: 10.1111/acer.70364 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Alcohol, Clinical and Experimental Research, 2026. Verfasst von Alexander S Hatoum und 9 weiteren Autor:innen. Beteiligte Einrichtung: Department of Psychiatry, Washington University School of Medicine, St. Louis, Missouri, USA. Hauptthemen laut MEDLINE: Alcoholism, Artificial Intelligence. Die Arbeit nennt 65 Literaturangaben. Der Volltext ist frei in PubMed Central verfügbar. Gefördert von: NIAAA NIH HHS, NIDA NIH HHS, National Institute on Alcohol Abuse and Alcoholism.

AbstractSchlagwörterBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Background

The wide‐ranging correlates (e.g., demographic, socioeconomic status [SES], brain, genetics) of alcohol use disorder (AUD) are typically characterized by small effects and studied in isolation. Novel multivariable nonlinear approaches may help characterize the multifaceted correlates of AUD and identify its strongest predictors.

Method

Here, using Gradient Boosted Machines (GBMs), we classify AUD in the UK Biobank ( N = 12,178) from features ( n = 440) within the following 5 domains: (1) demographic ( n = 12), (2) socioeconomic status ( n = 18), (3) non‐substance mental health ( n = 55), (4) non‐alcohol substance involvement ( n = 6), and (5) biological features ( n = 342). We use Explainable Artificial Intelligence (xAI) to characterize each respective feature's contribution to model performance and its heterogeneity, as well as the extent to which features from each domain collectively contribute to model performance.

Results

In an out‐of‐sample hold‐out test set ( n = 6089), our model classified lifetime AUD well (AUC = 0.80). Notably, feature importance analyses revealed that demographic (age, sex, income), substance use (ever smoke tobacco or ever use cannabis), and mental health questionnaires (social isolation) were most informative to classifying AUD. The gradient boosted machine significantly outperformed a linear model and implicated non‐zero pairwise interactions among variables, suggesting potential heterogeneity of AUD presentation by income and sex. Further, incrementally incorporating neuroimaging and genetic data as features led to small but consistent improvement in classifying AUD, with small effect sizes distributed across a number of variables (0%–2% gain based on overall model classification). This suggests the continued importance of biological features in classifying AUD beyond psychological and demographic measures, though classification with single selected brain regions may be limited.

Conclusion

Multimodal multivariable approaches considering nonlinear relationships may improve our understanding of AUD correlates. Particularly, accounting for interactions with demographic factors and integrating information across biological features may enhance model performance.

Schlagwörter

MeSH-Schlagwörter (MEDLINE)

AlcoholismArtificial IntelligenceHumansFemaleMaleMiddle AgedBoosting Machine Learning AlgorithmsUK BiobankClassification AlgorithmsAgedUnited Kingdom

Stichwörter der Autor:innen

alcohol use disorderexplainable AIgeneticmachine learningneuroimaging

Autor:innen und Institutionen

  • Department of Psychiatry, Washington University School of Medicine, St. Louis, Missouri, USA.
  • Department of Psychiatry, Indiana University School of Medicine, Bloomington, Indiana, USA.
  • Department of Psychological & Brain Sciences, Washington Univeristy in Saint Louis, St. Louis, Missouri, USA.
  • Department of Pharmacology and Toxicology, Medical College of Wisconsin, Milwaukee, Wisconsin, USA.

ORCID-Profile: Alexander S Hatoum, Alex P Miller, Hunter L Mathews, Aaron J Gorelik, Emma C Johnson

Förderung

  • NIAAA NIH HHS
  • NIDA NIH HHS
  • National Institute on Alcohol Abuse and Alcoholism
  • National Institute on Drug Abuse

Zitieren

APA

Hatoum, A. S., Miller, A. P., Mathews, H. L., Gorelik, A. J., Baranger, D. A. A., Luo, Z., Kutzer, J., Johnson, E. C., Agrawal, A., & Bogdan, R. (2026). Classification of Alcohol Use Disorder With Explainable AI. Alcohol, Clinical and Experimental Research. https://doi.org/10.1111/acer.70364

Vancouver

Hatoum AS, Miller AP, Mathews HL, Gorelik AJ, Baranger DAA, Luo Z, et al. Classification of Alcohol Use Disorder With Explainable AI. Alcohol, Clinical and Experimental Research. 2026. doi:10.1111/acer.70364

BibTeX
@article{Hatoum2026,
  author = {Hatoum, Alexander S and Miller, Alex P and Mathews, Hunter L and Gorelik, Aaron J and Baranger, David A A and Luo, Zhen and Kutzer, Jodi and Johnson, Emma C and Agrawal, Arpana and Bogdan, Ryan},
  title = {Classification of Alcohol Use Disorder With Explainable AI},
  journal = {Alcohol, Clinical and Experimental Research},
  year = {2026},
  doi = {10.1111/acer.70364}
}
Kostenlos starten →