EUVIMEDEUVIMED entdecken
Crossref

Leveraging Multimodal Self-Consistency Reasoning in Coding Motivational Interviewing for Alcohol Use Reduction

Guangzeng Han iD, James G Murphy iD, Benjamin O Ladd iD, Xiaolei Huang iD, Brian Borsari iD

Military Medicine · 01.08.2026 · 191 (Supplement_1) · 658-664

DOI: 10.1093/milmed/usag224 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Military Medicine, online erschienen am 06.08.2026. Verfasst von Guangzeng Han und 4 weiteren Autor:innen. Beteiligte Einrichtung: Department of Computer Science, University of Memphis, Memphis, TN 38152, United States. Hauptthemen laut MEDLINE: Motivational Interviewing, Alcohol Drinking. Die Arbeit nennt 20 Literaturangaben. Gefördert von: National Science Foundation.

AbstractSchlagwörterBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Introduction

Understanding client behaviors and predicting their outcomes require coding Motivational interviewing (MI) sessions with intensive labor costs and time consumptions of MI professionals. The advances in audio language models (ALMs) open promising opportunities in automating coding process, capturing critical multimodal signals of behavioral patterns, and interactive collaborations with expert annotators. Materials and Methods We experimented with 5 recorded sessions from de-identified MI audio tapes. We deployed audio-language models with four complementary analytic prompts to augment utterance-level reasoning: analytic (verbal cues), prosody-aware (acoustic cues), evidence-scoring (quantitative hypothesis test), and comparative (contrastive reasoning). Three stochastic samples were drawn per prompt, generating 12 independent reasoning trajectories per utterance, and majority votes of all trajectories decided predictions.

Results

Performance was evaluated using accuracy, precision, recall, and macro-F1 scores. The proposed MM-SC approach achieved 52.56% accuracy, 54.03% precision, 47.45% recall, and a macro-F1 score of 46.40%, exceeding the performance of baseline methods. Systematic ablation removing individual modules consistently degraded performance on primary metrics.

Conclusions

Multimodal self-consistency outperforms single-pass baseline prompting approaches. These findings suggest that incorporating both what clients say and how they say it can support more reliable MI coding.

Schlagwörter

MeSH-Schlagwörter (MEDLINE)

Motivational InterviewingAlcohol DrinkingHumans

Autor:innen und Institutionen

  • Department of Computer Science, University of Memphis, Memphis, TN 38152, United States.
  • Department of Psychology, University of Memphis, Memphis, TN 38152, United States.
  • Department of Psychology, Washington State University Vancouver, Vancouver, WA 98686, United States.
  • Integrative Healthcare Service, San Francisco Veterans Affairs Health Care System, San Francisco, CA 94121, United States.
  • Department of Psychiatry and Behavioral Sciences, University of California San Francisco, San Francisco, CA 94143, United States.

ORCID-Profile: Guangzeng Han, James G Murphy, Benjamin O Ladd, Xiaolei Huang, Brian Borsari

Förderung

  • National Science Foundation

Zitieren

APA

Han, G., Murphy, J. G., Ladd, B. O., Huang, X., & Borsari, B. (2026). Leveraging Multimodal Self-Consistency Reasoning in Coding Motivational Interviewing for Alcohol Use Reduction. Military Medicine, 191(Supplement_1), 658-664. https://doi.org/10.1093/milmed/usag224

Vancouver

Han G, Murphy JG, Ladd BO, Huang X, Borsari B. Leveraging Multimodal Self-Consistency Reasoning in Coding Motivational Interviewing for Alcohol Use Reduction. Military Medicine. 2026;191(Supplement_1):658-664. doi:10.1093/milmed/usag224

BibTeX
@article{Han2026,
  author = {Han, Guangzeng and Murphy, James G and Ladd, Benjamin O and Huang, Xiaolei and Borsari, Brian},
  title = {Leveraging Multimodal Self-Consistency Reasoning in Coding Motivational Interviewing for Alcohol Use Reduction},
  journal = {Military Medicine},
  year = {2026},
  volume = {191},
  number = {Supplement_1},
  pages = {658--664},
  doi = {10.1093/milmed/usag224}
}
Kostenlos starten →