Scholarlink

<p>In medicine, the laborious annotation process poses a significant limitation to the creation of large labelled datasets, a challenge that is particularly pronounced for medical video data, where numerous frames require expert annotation. Weakly supervised approaches, such as Multiple Instance Learning (MIL), offer a promising solution by leveraging only video-level labels rather than frame-by-frame annotations. In this work, we tailor a top-k MIL-based approach to medical video data to detect and localise facial motor tics in videos of patients diagnosed with Gilles de la Tourette Syndrome (GTS). Our adaptations include a score-driven top-k instance selection strategy that accounts for variability in tic occurrence between subjects. This enables the effective exploitation of large unlabelled datasets by incorporating weak labels where available. Evaluated against state-of-the-art MIL models, the proposed approach achieves the highest instance-level tic detection performance, with an area under the curve of 86.73 %. These results demonstrate that facial motor tics in videos of individuals with GTS can be effectively localised using only weak video-level annotations, highlighting the potential of the proposed approach for future clinical applications.</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

Weakly-Supervised Detection of Facial Motor Tics in Video Data Using Score-Driven Top-k Multiple Instance Learning

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2026年7月30日

查看全文
  • Nele Sophie Brügge,
  • Frédéric Li,
  • Ronja Schappert,
  • Gesine Marie Sallandt,
  • Christian Beste,
  • Sebastian Fudickar,
  • Marcin Grzegorzek,
  • Alexander Münchau,
  • Heinz Handels

摘要

In medicine, the laborious annotation process poses a significant limitation to the creation of large labelled datasets, a challenge that is particularly pronounced for medical video data, where numerous frames require expert annotation. Weakly supervised approaches, such as Multiple Instance Learning (MIL), offer a promising solution by leveraging only video-level labels rather than frame-by-frame annotations. In this work, we tailor a top-k MIL-based approach to medical video data to detect and localise facial motor tics in videos of patients diagnosed with Gilles de la Tourette Syndrome (GTS). Our adaptations include a score-driven top-k instance selection strategy that accounts for variability in tic occurrence between subjects. This enables the effective exploitation of large unlabelled datasets by incorporating weak labels where available. Evaluated against state-of-the-art MIL models, the proposed approach achieves the highest instance-level tic detection performance, with an area under the curve of 86.73 %. These results demonstrate that facial motor tics in videos of individuals with GTS can be effectively localised using only weak video-level annotations, highlighting the potential of the proposed approach for future clinical applications.