According to a new study presented at the European Academy of Dermatology and Venereology Congress 2023, artificial intelligence (AI) software can detect melanomas with 100% accuracy. This groundbreaking research highlights how AI is rapidly improving and learning, thanks to advancements in training techniques and data quality. Lead author Dr. Kashini Andrew, a specialist registrar at University Hospitals Birmingham NHS Foundation Trust, emphasized the significant role AI played in catching potentially fatal cancer cases in their early stages.
Over a period of two and a half years, the AI program analyzed 22,356 patients with suspected skin cancer. It successfully detected 99.5% of all skin cancers (189 out of 190), 92.5% of precancerous lesions (541 out of 585), and all 59 cases of melanoma. This marks a significant improvement compared to a previous AI model tested in 2021, which only detected 85.9% of melanoma cases, 83.8% of skin cancers, and 54.1% of precancerous lesions.
The latest version of the software has already made a notable impact by saving over 1,000 face-to-face consultations between April 2022 and January 2023, allowing more time for urgent patient cases. However, it is essential to note that the software should not replace the expertise of dermatologists but instead be used alongside professional care.
While the AI software demonstrated remarkable accuracy, researchers emphasize that it should not be solely relied upon for skin cancer detection. Co-author Dr. Irshad Zaki, a consultant dermatologist at UHB NHS Foundation Trust, stressed that AI should not be considered a substitute for dermatologists. A single case of basal cell carcinoma out of 190 was missed by the software, further highlighting the importance of having dermatologists provide a “safety net” for accurate diagnosis.
The role of AI in dermatology and its appropriate implementation are subjects of ongoing debate. Dr. Andrew recognized that further research, under appropriate clinical oversight, could lead to the use of AI as a triage tool. However, any implementation must demonstrate cost-effectiveness, and AI currently cannot function independently in dermatology. This study showcases the great promise AI holds for the future of healthcare provision while acknowledging its limitations.
Denial of responsibility! Vigour Times is an automatic aggregator of Global media. In each content, the hyperlink to the primary source is specified. All trademarks belong to their rightful owners, and all materials to their authors. For any complaint, please reach us at – [email protected]. We will take necessary action within 24 hours.
