A new deep learning algorithm can evaluate 134 different skin disorders, predicting malignancy and recommending key treatment options, according to new findings published in the Journal of Investigative Dermatology.
The AI model, trained using more than 220,000 images, was compared to the performance of practicing dermatologists, dermatology residents and members of the general public. The model achieved a performance comparable to dermatology residents but “slightly below” that of dermatologists.
The researchers also tracked how using the AI algorithm might impact a healthcare provider’s performance. The overall sensitivity of the malignancy diagnosis of participating dermatologists and dermatology residents, for instance, increased from 77.4% to 86.8%. The sensitivity of the malignancy diagnosis also saw an improvement, from 47.6% to 87.5%.
“Recently, there have been remarkable advances in the use of AI in medicine,” Jung-Im Na, MD, PhD, department of dermatology at Seoul National University in South Korea, said in a prepared statement. “For specific problems, such as distinguishing between melanoma and nevi, AI has shown results comparable to those of human dermatologists. However, for these systems to be practically useful, their performance needs to be tested in an environment similar to real practice, which requires not only classifying malignant versus benign lesion, but also distinguishing skin cancer from numerous other skin disorders including inflammatory and infectious conditions.”
It also turns out, according to Na, that reports of AI replacing human physicians may have been greatly exaggerated.
“Rather than AI replacing humans, we expect AI to support humans as ‘augmented intelligence’ to reach diagnoses faster and more accurately,” he said in the same statement.
Michael has more than 16 years of experience as a professional writer and editor. He has written at length about cardiology, radiology, artificial intelligence and other key healthcare topics.