Patients Appear Receptive to Use of Artificial Intelligence for Skin Cancer Screening

Patients appear open to the idea of using artificial intelligence (AI) to help in screening for skin cancer, however many patients stress the need to preserve the patient-physician relationship if implemented, according to study results published in JAMA Dermatology.

“In dermatology, researchers are evaluating the potential of machine learning to classify skin lesions using images from standard and dermoscopic cameras,” the researchers wrote. “Artificial intelligence may significantly alter how patients engage in health care, and the medical literature in this field is rapidly expanding. … However, our current understanding of how patients perceive AI and its application to health care lacks clarity and depth.”

To assess how patients conceptualize AI and perceive the use of AI for skin cancer screening, the researchers enrolled 48 patients from general dermatology clinics at the Brigham and Women’s Hospital and melanoma clinics at the Dana-Farber Cancer Institute from May 6, 2019 to July 8, 2019.

Analyzing the perceived risks and benefits, as well as the strengths and weaknesses, implementation, response to conflict between human and AI clinical decision making, and recommendation for or against the use of AI were the main goals of the study.

The participants were separated into three categories: A history of melanoma (16 patients), a history of nonmelanoma skin cancer (16 patients) and a history of no skin cancer (16 patients). Ninety percent of the participants reported owning at least one electronic device and 90% also reported using digital services for health.

Half of the patients in each cohort were interviewed about a direct-to-patient AI tool and the other half were interviewed about a clinician decision-support AI tool.

In interview responses, the most common perceived benefits patients cited for use of AI tools in skin cancer screening were increased diagnostic speed (60% of patients) and health care access (60% of patients).

Patients associated increased diagnostic speed with early skin cancer detection and lifesaving potential. The authors noted that one patient responded that AI “could reach people who don’t have great access to health care but may have an iPhone.” Additional perceived benefits the patients cited included reduced health care cost (35%), reduced patient anxiety (33%) and increased triage efficiency (29%).

Forty percent pf patients perceived the greatest risk of AI use for skin cancer screening was increased patient anxiety. Additional common perceived risks included the loss of human social interaction (38%), the loss of privacy (29%) and patient loss to follow-up (29%).

One of the strengths of AI compared with skin cancer screening conducted by a human was a patient’s perceived more accurate diagnosis (69%). That perception, according to the authors, was based on the ability of the AI to learn, evolve and share data, as well as have more experience than humans.

Regardless of perceived strengths and weaknesses from patients, most responded that it was important to have a relationship between humans and the AI (94%). The authors report that patients envisioned the AI referring to a physician and providing a second opinion.

In an event there was conflicting diagnoses between human and AI, the most common response from patients was to seek a biopsy (67%). Patients also responded that they would either put more faith in their doctor (60%) or seek a second opinion form another physician (42%).

Most patients (75%) responded that they would recommend the AI to family and friends. However, 19% were ambivalent to the process and 6% would not recommend it at all.

“Our results indicate that most patients are receptive to the use of AI for skin cancer screening within the framework of human-AI symbiosis,” the authors concluded. “Although additional research is required, the themes that emerged in this study have important implications across the house of medicine. Through patients’ eyes, augmented intelligence may improve health care quality but should be implemented in a manner that preserves the integrity of the human physician-patient relationship.”