Real-world diagnostic performance of artificial intelligence in skin cancer screening
Charles H Earnshaw, Karmen Cheung, Fasiha Arshad, Ross Moore, Ali Al-Janabi, Emma McMullen · 2026
Read Real-world diagnostic performance of artificial intelligence in skin cancer screening by Charles H Earnshaw, Karmen Cheung, Fasiha Arshad, Ross Moore, Ali Al-Janabi, Emma McMullen (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/pmc-13425094.txt.
Description
Earnshaw and colleagues prospectively evaluated an artificial intelligence system, DERM, during its first three months inside a National Health Service urgent skin cancer referral pathway in North West England. They compared the algorithm's verdicts with face-to-face dermatologist diagnoses and with the final tissue diagnosis after biopsy.
The AI reached a sensitivity of 95.3 percent for skin cancer, significantly higher than the dermatologists' 88.5 percent in the same pathway. Its positive predictive value was lower, however, at 46.5 percent against 62.1 percent for dermatologists judging the same lesions. Across 318 assessments that had no remote human review, the algorithm named the exact diagnosis only 28.6 percent of the time, versus 61.6 percent for dermatologists. Four cancers were found among lesions the AI had classed as benign, prompting caution about removing human review.
Catalog record
| CREATOR | Charles H Earnshaw, Karmen Cheung, Fasiha Arshad, Ross Moore, Ali Al-Janabi, Emma McMullen |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 28 KB |
| VERSION | 1.0 |
| LICENSE | CC BY 4.0 |
| RIGHTS | CC BY |
| SOURCE | Europe PMC |