Evaluation of a Voice-Based Emotion Recognition Software in the Psycho-Oncological Care of Cancer Patients

Abstract

Due to the increasing global incidence of cancer, the growing number of long-term survivors and the prevalence of psychological distress among cancer patients, psycho-oncological support is becoming more crucial. Recognizing the rising demand for psycho-oncological care the “Cancer Counselling App” project was initiated. As part of this project, a cancer counselling app is being developed. The development of the app incorporates the investigation a of voice-based emotion recognition which is enabled through the increasing capabilities of machine and deep learning algorithms, aiming to support the psycho-oncological care of cancer patients. The objective of this study is to identify use cases for this functionality and determine which of them are suitable for enhancing the psycho-oncological care. Through a literature review and expert interviews, seven distinct use cases were identified and evaluated. The highest-priority use case for voice-based emotion recognition is the long-term monitoring of the emotional state of cancer patients. The functionality should particularly focus on the emotions anxiety and distress, along with the psychological disorder depression, to effectively support psycho-oncological treatment.