Confirming that you would like me to rewrite the provided text with almost no loss of volume. Here is the rewritten version: Please confirm your intention to log out of Medscape. Logging out will require you to enter your username and password during your next visit. A recent study conducted by UK researchers reveals that a new artificial intelligence (AI) model can accurately detect the most lethal form of skin cancer with a 100% success rate. The AI was able to detect over 99% of all skin cancers examined. During the testing phase, the researchers integrated the AI into a clinical diagnosis process, suggesting a future where AI assists doctors in detecting skin cancer more efficiently and prioritizing patients. Skin cancer is the most prevalent type of cancer in the US, affecting 1 in 5 Americans by the age of 70. The five-year survival rate for melanoma, the most dangerous form of skin cancer, exceeds 99% if detected early, but only approximately 75% of melanomas are caught at this stage. Concerns have arisen regarding the insufficient number of dermatologists in the workforce to meet the escalating rates of skin cancer. Consequently, the average wait time for a dermatology appointment has increased and reached 34. 5 days in 2022. The study, presented at the European Academy of Dermatology and Venereology Congress, involved 6, 900 patients in the UK who were referred by their primary care physicians due to suspected skin cancer. Images of the suspicious areas were captured and uploaded to the AI software. The software's assessment was then shared with a dermatologist. Lead researcher Dr. Kashini Andrew, a dermatologist and specialist registrar at University Hospitals Birmingham NHS Foundation Trust, clarified that the dermatologist performing the second assessment was aware of the AI's diagnosis. The AI successfully identified 59 out of 59 melanoma cases and 189 out of 190 total skin cancer cases (99. 5%). It also flagged 541 out of 585 precancerous lesions (92. 5%).

These results demonstrate significant improvement compared to a previous version of the model developed in 2021, which detected 86% of melanomas, 84% of all skin cancers, and 54% of precancerous lesions. Over the course of the 10-month study, the AI system saved more than 1, 000 in-person consultations, allowing dermatologists to focus their efforts on detecting more cancers and serving more patients. It's worth noting that the study's sample size was limited to patients from a single hospital in one region of the UK, which prevents making broad generalizations about the use of AI in dermatology. However, it encourages further discussion. Dr. Roxana Daneshjou, a dermatologist at Stanford who has extensively studied the advantages and disadvantages of AI in medicine, expressed some concerns. One concern is that doctors can obtain more detailed information during a physical examination than AI can deduce from a photo. Doctors can examine the texture of the skin, gather patient history, and take photos using specialized lighting and magnification. Another area for improvement lies in the AI's ability to confidently rule out malignancy. In this study, the AI identified 75% of benign lesions, indicating a decline compared to the previous version. The researchers acknowledged the potential trade-off between increased sensitivity and the risk of unnecessary biopsies, which can burden the healthcare system, incur costs, and cause stress and scarring. Nevertheless, if AI software similar to the one used in this study proves equally accurate in larger and more diverse sample sizes, it could become a valuable tool for triaging patients. Dr. Daneshjou suggests that if AI becomes adept at both identifying and ruling out malignancy, it would be a significant achievement. To stay updated with the latest news, follow Medscape on Facebook, X, Instagram, and YouTube. You have successfully subscribed to receive email alerts when new content is published.