Scholarlink

<p>Music therapy, as a non-pharmaceutical intervention in evidence-based medicine, requires systematic knowledge of clinically effective musical content. However, existing datasets are rarely purpose-built for clinical music therapy and often lack fine-grained clinical annotations, limiting personalized music therapy guidance. We introduce MCUD 1.0, a clinical music usage data resource comprising structured records for 42 playlists and 378 songs, without distributing audio recordings. The resource was curated from randomized controlled trials indexed in PubMed (2002–2024) and spans diverse clinical contexts, including neurodegenerative disease, affective and anxiety disorders, cognitive impairment, and cardiovascular conditions. It provides playlist-level annotations (treatment goal, disease information, medical procedure, and intervention modality), song-level metadata and annotations (genre, emotion, and lyrics), curated source links, and traditional acoustic features and processing code for generating deep audio embeddings and mel-spectrogram images. We assessed data completeness across the released data files. We conducted baseline machine-learning validation on three tasks, including clinical usage prediction, emotion recognition, and genre classification, achieving best micro F1 scores of 0.623, 0.271, and 0.510, respectively. We examined acoustic feature and label associations and observed interpretable relationships linking rhythmic, spectral, and tonal descriptors with clinical usage, emotion, and genre annotations. MCUD 1.0 provides a high-quality and traceable clinical data resource that will be iteratively updated, supporting the development of intelligent systems for personalized music therapy.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

MCUD 1.0: A Clinical Music Usage Data Resource for Personalized Music Therapy

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2026年7月28日

查看全文
  • Jiale Du,
  • Jie Song,
  • Zhichuan Xu,
  • Yuan Fang,
  • Yuxin Zhang,
  • Cheng Bi,
  • Xin Zheng,
  • Bairong Shen

摘要

Music therapy, as a non-pharmaceutical intervention in evidence-based medicine, requires systematic knowledge of clinically effective musical content. However, existing datasets are rarely purpose-built for clinical music therapy and often lack fine-grained clinical annotations, limiting personalized music therapy guidance. We introduce MCUD 1.0, a clinical music usage data resource comprising structured records for 42 playlists and 378 songs, without distributing audio recordings. The resource was curated from randomized controlled trials indexed in PubMed (2002–2024) and spans diverse clinical contexts, including neurodegenerative disease, affective and anxiety disorders, cognitive impairment, and cardiovascular conditions. It provides playlist-level annotations (treatment goal, disease information, medical procedure, and intervention modality), song-level metadata and annotations (genre, emotion, and lyrics), curated source links, and traditional acoustic features and processing code for generating deep audio embeddings and mel-spectrogram images. We assessed data completeness across the released data files. We conducted baseline machine-learning validation on three tasks, including clinical usage prediction, emotion recognition, and genre classification, achieving best micro F1 scores of 0.623, 0.271, and 0.510, respectively. We examined acoustic feature and label associations and observed interpretable relationships linking rhythmic, spectral, and tonal descriptors with clinical usage, emotion, and genre annotations. MCUD 1.0 provides a high-quality and traceable clinical data resource that will be iteratively updated, supporting the development of intelligent systems for personalized music therapy.