Nicholas Doherty
Research Mentor(s): Sekhar Sripada
Mentor Department: Psychiatry/Philosophy
Authors: Nicholas Doherty, Sekhar Sripada
Session: Session 6 (3:00pm – 3:50pm)
Presentation Type: Poster 39
Abstract
Psychiatric clinical care appointments begin with the clinician asking the patient to recap or describe how their past week or month went. This process carries many different flaws because of patients’ habit to focus too much on recent events and many times find it difficult to pinpoint how they felt at a specific time. This makes appointments take longer and makes the data gathering and decision process of the psychiatric clinician more difficult. The solution our team is working on is the use of voice diaries recorded each night by the patient to gather more accurate data. After the voice diary data is gathered, generative AI will transcribe the voice diaries and summarize them into graphs and mood summaries. The project is still in the process of being worked on with our current task being to validate the responses given to us by the generative AI with the final step and main component being students grading the responses based on accuracy and informativeness.The development and implementation of voice diaries combined with generative AI transcription and summarization has significant implications for psychiatric clinical care. By addressing the issue of recency bias and recall challenges in patient self-reports, this approach has the potential to provide clinicians with more accurate and comprehensive insights into a patient’s emotional state over time. This can lead to more informed and efficient decision-making processes, ultimately improving the quality and effectiveness of psychiatric treatment. It can also work as an initial use of generative AI in the medical field opening the possibilities for further use and research on how generative AI methods could be applied.
