• Case-Based Roundtable
  • General Dermatology
  • Eczema
  • Chronic Hand Eczema
  • Alopecia
  • Aesthetics
  • Vitiligo
  • COVID-19
  • Actinic Keratosis
  • Precision Medicine and Biologics
  • Rare Disease
  • Wound Care
  • Rosacea
  • Psoriasis
  • Psoriatic Arthritis
  • Atopic Dermatitis
  • Melasma
  • NP and PA
  • Skin Cancer
  • Hidradenitis Suppurativa
  • Drug Watch
  • Pigmentary Disorders
  • Acne
  • Pediatric Dermatology
  • Practice Management
  • Prurigo Nodularis
  • Buy-and-Bill

Article

Artificial Intelligence and Skin Cancer

Laura Ferris, MD, PhD, discusses the future landscape of artificial intelligence and skin cancer detection at SBS 2023.

Artificial intelligence (AI), or augmented intelligence, can be applied to any image of lesions that are suspicious of skin cancer or even inflammatory lesions, according to Laura Ferris, MD, PhD, associate professor of dermatology and clinical and translational science at the University of Pittsburgh, and director of the clinical trials office at the University of Pittsburgh Medical Center in Pittsburgh, Pennsylvania. Ferris discussed how AI can be used to detect skin cancer in her South Beach Symposium (SBS) 2023 presentation, “Artificial Intelligence and Detection.”

“The basic principle of AI is that you have a training set where you have a known diagnosis, you train a computer; we can train it on features we know, or we can sort of let the computer identify and pick the features that it has found to be good differentiators. And then we test this on a new set of new images,” said Ferris.

In her discussion, Ferris dove into the difference between AI and human detection. A machine can only analyze what a dermatologist instructs it to look at. Dermatologists have other factors that they can assess, such as patient history, skin lesions, and the patient’s entire body in general. AI is limited to what dermatologists choose to show it, or even what a patient chooses to show.

AI is objective. Ferris noted that “machines are not going to lose sleep over missing skin cancers like we do.” Dermatologists also typically favor biopsying benign lesions rather than missing cancer, but AI will only do what it is told to based on the thresholds set by the dermatologist.

Another key point of Ferris’ presentation was smartphone apps that are marketed for detecting skin cancer. Ferris completed a previous study that compared the accuracy of skin cancer diagnoses from skin cancer detection apps. The data demonstrated that the sensitivity of detection was quite low, but these apps continue to be available for the public to use.

One issue that Ferris emphasized is that these skin cancer detection apps are not regulated by the US Food and Drug Administration (FDA). The FDA does not have oversight of AI technology in general. There continues to be an ongoing discussion of how to validate a skin cancer detection app. Ferris mentioned that the FDA has said that they believe that an app that analyzes a skin lesion and provides a user with an assessment of risk is something that they should oversee.

Reference

Ferris L. Artificial intelligence and detection. Presented at the 2023 South Beach Symposium Meeting; February 9-12, 2023; Miami Beach, FL.

Related Videos
© 2024 MJH Life Sciences

All rights reserved.