Creating a Secure Audiovisual Database of Adults With Cancer to Develop AI Tools for Pain Assessment

What this Trial Is About

Researchers are conducting an observational study to build a secure, high-quality audiovisual database of adults with cancer to aid future artificial intelligence AI research in pain assessment. The study aims to capture how people with cancer express pain and non-pain experiences through short video and sound recordings. This research includes two groups those admitted for cancer-related pain and those admitted for chemotherapy without pain. Participants will be video recorded from the shoulders up for up to 60 seconds while reading a short sentence and describing their pain or daily experience. They will complete a brief questionnaire about their mood and pain expression. Those in the pain group may have recordings on several consecutive days during their admission, while the control group will be recorded once. Clinical information such as pain scores, medications, and cancer type will be collected from medical records. No experimental treatments or extra hospital visits are involved. During the study, participants will be assessed through audiovisual recordings and questionnaires. Researchers will collect clinical data from electronic health records and analyze facial and vocal features using open-source software. The primary outcome is the audiovisual recordings of facial and vocal pain expression collected up to two weeks. This study provides a foundation for developing AI tools to support more accurate pain monitoring in clinical care, with all data securely stored and handled with privacy protections.

Developing a Multimodal Cancer Pain Database to Support AI-Based Automatic Pain Assessment