Depression Diagnosis Using Multimodal Emotion Data with Modality Attention Networks

2023 Fiscal Year Final Research Report

• Project/Area Number: 22K21316
• Research Category: Grant-in-Aid for Research Activity Start-up
• Allocation Type: Multi-year Fund
• Review Section: 1002:Human informatics, applied informatics and related fields
• Research Institution: Ritsumeikan University
• Principal Investigator: Liu Jiaqing 立命館大学, 情報理工学部, 助教 (20948343)
• Project Period (FY): 2022-08-31 – 2024-03-31
• Keywords: マルチモダリティAI / うつ状態自動認識 / 情動データベース / 動的歩行

Outline of Final Research Achievements

This study aimed to enhance the accuracy of depression diagnosis by integrating multimodal information such as facial expressions, voice, and gait. Utilizing the Modality Attention Network, a new method was proposed to effectively fuse information from various modalities, achieving more accurate recognition of depressive states. The findings contribute to improving the precision of clinical diagnoses and are expected to have a significant social impact by enhancing the quality of depression treatment.

Free Research Field

人工知能

Academic Significance and Societal Importance of the Research Achievements

本研究は、モダリティアテンションネットワークを用いて、うつ病の正確な診断と早期発見を可能にする新しいアプローチを提案した。この方法は、複数のモダリティからの情報を統合し、診断プロセスの信頼性と精度を向上させる。この研究の成果は、医療診断技術の進展に寄与し、うつ病治療の効率化と患者の生活質の改善につながる社会的意義も持つ。