Credit score: Unsplash/CC0 Public Area
A easy AI mannequin has been proven to carry out on a par with skilled dermatologists when assessing the aggressiveness of a standard type of pores and skin most cancers, squamous cell carcinoma. The analysis was headed by the College of Gothenburg.
Annually, greater than 10,000 Swedes develop squamous cell carcinoma. That is the second commonest type of pores and skin most cancers in Sweden, after basal cell carcinoma, and its prevalence is growing quickly. Squamous cell carcinoma usually develops within the head and neck area and different areas uncovered to the solar over a few years.
“This type of cancer, which is a result of mutations of the most common cell type in the top layer of the skin, is strongly linked to accumulated UV radiation over time. It develops in sun-exposed areas, often on skin already showing signs of sun damage, with rough scaly patches, uneven pigmentation, and decreased elasticity,” says affiliate professor and dermatologist Sam Polesie, who led the research.
Squamous cell carcinoma analysis is commonly simple—the problem lies within the preoperative evaluation—figuring out how aggressively the tumor is rising to plan and prioritize surgical procedure appropriately. If the tumor is extra aggressive, the surgical procedure must be scheduled promptly, with extra adjoining tissue eliminated. For much less aggressive tumors, narrower margins can be utilized, with easier procedures adequate in some instances.
Nearly an identical efficiency
In lots of international locations, Sweden included, preoperative punch biopsies usually are not routinely carried out for suspected squamous cell carcinoma. Surgical procedure is as an alternative carried out primarily based solely on the scientific suspicion of a tumor, with the complete excised specimen despatched for histopathological evaluation.
The truth that surgical procedure is carried out with no preoperative biopsy underscores the necessity for evaluation options that don’t require tissue samples, equivalent to picture evaluation utilizing synthetic intelligence (AI).
For the research, the researchers educated an AI system in picture evaluation utilizing 1,829 scientific close-up pictures of confirmed squamous cell carcinoma. The AI mannequin’s potential to tell apart three ranges of tumor aggressiveness was then examined on 300 pictures and in contrast with the assessments of seven impartial skilled dermatologists.
The outcomes, printed within the Journal of the American Academy of Dermatology Worldwide, present that the AI mannequin carried out nearly identically to the workforce of medical specialists. On the identical time, settlement between particular person dermatologist assessments was solely average, underscoring the complexity of the duty.
Two scientific options—ulcerated and flat pores and skin surfaces—have been discovered to be clearly related to extra aggressive tumor development. Tumors exhibiting these traits have been greater than twice as more likely to fall into one of many two larger ranges of aggressiveness.
Well being care wants ought to resolve
The usage of synthetic intelligence in pores and skin most cancers care has attracted a substantial amount of curiosity in recent times, though based on Polesie, up to now it has had restricted sensible affect on well being care. He emphasizes the significance of clearly outlined software areas the place analysis can create added worth for Swedish well being care.
“We believe that one such application area could be the preoperative assessment of suspected skin cancers, where more nuanced conclusions can influence decisions. The model we’ve developed needs further refinement and testing, but the way forward is clear—AI should be integrated where it actually adds value to decision-making processes within health care,” says Polesie.
Polesie is an affiliate professor of dermatology and venereology on the College of Gothenburg and a practising dermatologist at Sahlgrenska College Hospital. The photographs comprising the research knowledge have been taken inside dermatological well being care on the college hospital between 2015 and 2023.
Extra info:
Victor Liang et al, Assessing differentiation in cutaneous squamous cell carcinoma: A machine studying method, JAAD Worldwide (2025). DOI: 10.1016/j.jdin.2025.07.004
Offered by
College of Gothenburg
Quotation:
AI matches dermatologists in assessing frequent pores and skin most cancers (2025, September 12)
retrieved 13 September 2025
from https://medicalxpress.com/information/2025-09-ai-dermatologists-common-skin-cancer.html
This doc is topic to copyright. Other than any honest dealing for the aim of personal research or analysis, no
half could also be reproduced with out the written permission. The content material is offered for info functions solely.
