EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

MIDAS: A multimodal interview-based framework for depression assessment from clinical interviews

Yuyang Sha, Ting Liu, Xianyu Tong, Gang Luo, Xiangyu Guo, Jiasong Xiao, Zitao Liu, Wei Chen, Yipeng Wang

Acta Psychologica · 2026-10 · 270 · 107700

DOI: 10.1016/j.actpsy.2026.107700 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Acta Psychologica, 2026-10. Verfasst von Yuyang Sha und 8 weiteren Autor:innen. Beteiligte Einrichtung: Eye Research Center, Hangzhou Institute of Medicine, Chinese Academy of Sciences, Eye Hospital, Wenzhou Medical University, Hangzhou, 310018, China. …. Hauptthemen laut MEDLINE: Depression, Interview, Psychological. Die Arbeit nennt 46 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: National Natural Science Foundation of China.

AbstractSchlagwörterBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Automated depression assessment from clinical interviews is a challenging intelligent healthcare task because depressive states are reflected not only in verbal responses, but also in acoustic and visual-behavioral cues. Existing multimodal methods have improved prediction performance, but many still process interviews as generic long sequences and provide limited structured evidence for interpreting model outputs. This paper presents MIDAS, an interview-aware multimodal framework that preserves the question-response structure of clinical interviews for depression assessment. MIDAS organizes interviews into question-response pairs to preserve the interaction structure of clinical interviews, and integrates linguistic, acoustic, and visual-behavioral representations through gated multimodal fusion and shared large language model-based contextual modeling. The framework jointly supports depression classification, PHQ-8 severity prediction, and structured multimodal evidence summarization. Experiments on the DAIC-WOZ development set show that MIDAS achieves a Macro-F1 of 88.10% for depression classification, with Dep-F1 and Con-F1 scores of 85.76% and 90.45%, respectively. For PHQ-8 severity prediction, MIDAS achieves an MAE of 3.03 and an RMSE of 3.93. Ablation results indicate that acoustic and visual-behavioral information provide complementary benefits over linguistic-only modeling. Qualitative examples further show that MIDAS can organize modality-specific behavioral cues into structured evidence summaries associated with model predictions. These results demonstrate the potential of interview-aware multimodal modeling as a research framework for depression assessment, while the structured evidence summaries are intended to support the interpretation of model outputs.

Schlagwörter

MeSH-Schlagwörter (MEDLINE)

DepressionInterview, PsychologicalHumansLarge Language Models

Stichwörter der Autor:innen

Behavioral signal analysisClinical interview analysisDepression assessmentExplainable artificial intelligenceLarge language modelsMultimodal learning

Autor:innen und Institutionen

  • Eye Research Center, Hangzhou Institute of Medicine, Chinese Academy of Sciences, Eye Hospital, Wenzhou Medical University, Hangzhou, 310018, China. Electronic address: shayuyang@eye.ac.cn.
  • MT Lab, Meitu Inc., Beijing, 100191, China.
  • National Cancer Center/National Clinical Research Center for Cancer/Cancer Hospital, Chinese Academy of Medical Sciences & Peking Union Medical College, Beijing, 100730, China.
  • Center for Artificial Intelligence Driven Drug Discovery, Faculty of Applied Sciences, Macao Polytechnic University, 999078, Macao Special Administrative Region of China.
  • Eye Research Center, Hangzhou Institute of Medicine, Chinese Academy of Sciences, Eye Hospital, Wenzhou Medical University, Hangzhou, 310018, China.
  • Cixi Biomedical Research Institute, Wenzhou Medical University, Wenzhou, 325027, China; National Clinical Research Center for Ocular Diseases, Eye Hospital, Wenzhou Medical University, Wenzhou, 325027, China; State Key…

Förderung

  • National Natural Science Foundation of China

Zitieren

APA

Sha, Y., Liu, T., Tong, X., Luo, G., Guo, X., Xiao, J., Liu, Z., Chen, W., & Wang, Y. (2026). MIDAS: A multimodal interview-based framework for depression assessment from clinical interviews. Acta Psychologica, 270, 107700. https://doi.org/10.1016/j.actpsy.2026.107700

Vancouver

Sha Y, Liu T, Tong X, Luo G, Guo X, Xiao J, et al. MIDAS: A multimodal interview-based framework for depression assessment from clinical interviews. Acta Psychologica. 2026;270:107700. doi:10.1016/j.actpsy.2026.107700

BibTeX
@article{Sha2026,
  author = {Sha, Yuyang and Liu, Ting and Tong, Xianyu and Luo, Gang and Guo, Xiangyu and Xiao, Jiasong and Liu, Zitao and Chen, Wei and Wang, Yipeng},
  title = {MIDAS: A multimodal interview-based framework for depression assessment from clinical interviews},
  journal = {Acta Psychologica},
  year = {2026},
  volume = {270},
  pages = {107700},
  doi = {10.1016/j.actpsy.2026.107700}
}
Kostenlos starten →