MelaTools-ML qualitative interview study

This study aimed to understand the views of stakeholders about using AI technologies to help detect skin cancer. We spoke to 29 patients, members of the public, general practitioners, primary care nurse practitioners, dermatologists, and AI researchers from universities and commercial companies. The interviews were held online between September 2022 and May 2023 using semi-structured qualitative interviews. The interviews were analyzed using thematic framework analysis.

The most common theme that was discussed in the interviews was about the best way for AI to be used, whether giving the AI technology to patients was the best option, or to primary care nurses, or to GPs and other primary care clinicians, or several other alternatives that were suggested. There was no clear agreement about who was best positioned to use the AI technology, but most people wanted to see the technology in the hands of either patients or GPs and primary care practitioners. We also discussed what the aim of the AI should be, there were lots of suggestions such as to rule out skin cancer, to diagnose skin lesions, to help clinicians make better referrals, to help patients take better photos of their skin lesions.

There were several themes in the interviews that cut across all aspects of AI development, testing, implementation and use. These included topics like trust, ease of use, acceptability, whether the AI works equally well for the whole population or excludes some patient groups, testing and regulation, implementation, and whether the AI technology can be used long-term. Participants were concerned about how good the data used to develop and test AI technologies is and how this affects their accuracy, particularly in patients with darker skin tones, and whether it would risk missing skin cancers or create bias. How easy the AI is to use and that it doesn’t increase the workload of healthcare services were both important for participants. Health care professionals and AI researchers worried that there aren’t established ways of testing and regulating AI technologies.

This study is one of the first to speak to stakeholders and gather their views on the use of AI technologies to help detect skin cancers. The best way to use these technologies has not yet been shown, but it is clear that AI technologies need to be developed, tested, and implemented carefully and thoughtfully if they are to be a useful tool. Care needs to be taken to make sure AI technologies are accurate with patients from all demographics and that they do not exclude any patient groups.

Read the full publication: Jones et al. User and Developer Views on Using AI Technologies to Facilitate the Early Detection of Skin Cancers in Primary Care Settings: Qualitative Semistructured Interview Study. JMIR Cancer 2025;11:e60653. doi: 10.2196/60653