Slideshow: AI in Skin Cancer Awareness Creates New Opportunities for Care

News
Slideshow

Research exploring the growing role of artificial intelligence (AI) in health care sheds a light on the potential of the technology to improve skin cancer detection and awareness among patients and health care professionals.

Earlier this year, the FDA cleared the first-of-its-kind DermaSensor, a real-time, non-invasive skin cancer evaluation system, for use in non-dermatology primary care settings.1 The handheld device uses painless spectroscopy technology and artificial intelligence (AI) analysis of suspicious lesions to provide feedback about skin cancer risks in patients.

Previously limited to visual mole assessments, primary care physicians are now empowered by DermaSensor, a quantitative, point-of-care tool, to provide precise and timely skin care evaluations. This innovation has the potential to relieve burdens associated with a shortage of dermatologists and accelerate necessary patient care.

The clearance of DermaSensor highlights the growing role of AI in health care. Although use of the device has not yet trickled into the pharmacy setting, other tools like automated dispensing systems and predictive analytics have already begun digitizing the field.

The rapid pace of AI development highlights a crucial opportunity for pharmacists. As the most accessible health care providers—and often the first to be approached by patients with undiagnosed skin conditions—they are well-positioned to learn about evolving AI applications in dermatology. This knowledge can empower the health care professionals to leverage AI’s potential in future care.

slideshow image

Using AI to assist human specialists in diagnosing skin cancer has the potential to improve diagnostic accuracy among medical professionals on all experience levels, according to results of a meta-analysis and systematic review published in Nature.2

Although AI has previously demonstrated its ability to accurately diagnose skin cancers, its dependence on specific datasets can lead to inconsistencies if the real-world data it encounters differs significantly than its input. To address this, researchers explored how AI and human specialists can work together to reinforce AI’s diagnoses and overcome limitations of human clinicians.

Twelve studies including 67,700 diagnostic evaluations of potential skin cancer were selected across PubMed, Embase, IEE Xplor, Scopus, and other conference proceedings published between January 1, 2017, and November 8, 2022, for assessment in the final meta-analysis and systematic review. To be eligible for inclusion, studies must have compared the diagnostic performance of clinicians on at least 1 skin cancer case, with and without the aid of deep learning-based AI assistance.

Among the selected studies, outpatient clinical images (42%), followed by dermoscopic images (33%), and in-patient visits (25%), were most frequently leveraged for diagnosis. Human doctors and AI assistance were tasked with either choosing the most likely diagnosis (58%) or categorizing the lesion as malignant or benign (42%).

Clinicians without AI assistance scored 74.8% (95% CI 68.6-80.1) on overall sensitivity and 81.5% (95% CI 73.9-87.3) on specificity. With AI assistance, overall sensitivity and specificity increased to 81.1% (95% CI 74.4-86.5) and 86.1% (95% CI 79.2-90.9), respectively, for clinicians on every experience level.

Further subgroup analyses revealed the greatest improvement in diagnostic accuracy with AI assistance occurred among non-dermatologists, such as primary care physicians, nurse practitioners, and medical students. This group experienced a 13-point increase in sensitivity and an 11-point increase in specificity with AI assistance.

Although their research warrants further exploration, investigators noted that their findings suggest health care professionals with the least experience in dermatology may benefit the most from AI assistance in building knowledge, making referrals, and counseling patients about skin cancer. This points to potential benefits in settings such as pharmacies, where staff may have less specialized training in the field.

READ MORE: Dermatology Resource Center

Pharmacy practice is always changing. Stay ahead of the curve: Sign up for our free Drug Topics newsletter and get the latest drug information, industry trends, and patient care tips, straight to your inbox.

References
1. FDA clearance granted for first AI-powered medical device to detect all three common skin cancers (melanoma, basal cell carcinoma, and squamous cell carcinoma). News release. DermaSensor Inc. January 17, 2024. Accessed May 23, 2024. https://www.businesswire.com/news/home/20240117116417/en/
2. Krakowski I, Kim J, Cai ZR, et al. Human-AI interaction in skin cancer diagnosis: a systematic review and meta-analysis. NPJ Digit Med. 2024;7(1):78. Published 2024 Apr 9. doi:10.1038/s41746-024-01031-w
3. Roffman D, Hart G, Girardi M, Ko CJ, Deng J. Predicting non-melanoma skin cancer via a multi-parameterized artificial neural network. Sci Rep. 2018;8(1):1701. Published 2018 Jan 26. doi:10.1038/s41598-018-19907-9
4. Key statistics for basal and squamous cell skin cancers. Data sheet. American Cancer Society. October 31, 2023. Accessed May 28, 2024. https://www.cancer.org/cancer/types/basal-and-squamous-cell-skin-cancer/about/key-statistics.html
5. Nelson CA, Pérez-Chada LM, Creadore A, et al. Patient perspectives on the use of artificial intelligence for skin cancer screening: A qualitative study. JAMA Dermatol. 2020;156(5):501-512. doi:10.1001/jamadermatol.2019.5014
Related Content
© 2025 MJH Life Sciences

All rights reserved.