# Cancer Patients' Perception, Acceptance, and Utilization of Artificial Intelligence‐Based Emotional Distress Assessment Tools: A Scoping Review

## TL;DR

This scoping review explores cancer patients' views on AI tools that assess emotional distress through voice, speech, and facial expressions.

## Contribution

The study is the first to synthesize cancer patients' perceptions and acceptance of AI-based emotional distress screening tools.

## Key findings

- High acceptance and satisfaction rates (70%-98%) were reported for AI-based distress screening tools.
- Three studies showed favorable views of AI tools for detecting emotional distress in cancer patients.
- Facial expression and speech semantics technologies were primarily used in the reviewed studies.

## Abstract

Emotional distress in cancer patients and survivors impacts overall well‐being and quality of life. Several barriers to adequate screening have been identified and are currently being addressed by artificial intelligence (AI)‐based tools. However, there is a critical need to explore cancer patients' and survivors' perspectives on these new technologies. This scoping review aims to synthesize the available evidence on their perception, acceptance, and utilization of AI‐based voice, speech semantics, and facial expression (AIVSFE) tools for emotional distress screening.

A systematic search was conducted in Scopus, Web of Science, PubMed Central, Cochrane Central Register of Controlled Trials (CENTRAL), PsycINFO, and Epistemonikos on July 1, 2025. Empirical studies published from January 1, 2019, to the search date that focused on adult cancer patients at any stage of treatment or survivorship and their perception, acceptance, or use of AIVSFE tools were retrieved. Participant sociodemographics, AI‐based distress screening modalities, technological frameworks, measurement tools, outcomes, and the studies' methodological quality were analyzed.

Three studies met the eligibility criteria. They included a combined sample of 316 cancer patients and survivors with heterogeneous clinical characteristics. Two studies utilized speech semantics technologies, while one utilized facial expression technology. The results show high acceptance, satisfaction, and usefulness rates (70%–98%), suggesting AIVSFE tools could address barriers associated with traditional distress screening.

The findings indicate a favorable view of AIVSFE tools for detecting distress. Future studies should prioritize developing standardized evaluation frameworks, diversifying participant demographics, and addressing broader usability and ethical concerns to ensure equitable adoption of these technologies.