Medicine Monday: Harnessing AI to diagnose skin cancerDec 21, 2020
Melanoma is the deadliest form of skin cancer, but it is highly curable if diagnosed and treated early. That is why dermatologists and researchers like Dr. Allan C. Halpern at Memorial Sloan Kettering Cancer Center (MSKCC) are working to find new, more accurate ways to detect skin cancer in patients.
The growing power of artificial intelligence (AI), combined with the rapid adoption of telemedicine due to the COVID-19 pandemic, presents both opportunities and risks for skin cancer care. To better understand these implications, Dr. Halpern and a team of researchers from over 16 medical institutions worldwide devised a study to assess the accuracy of image-based AI for cancers like melanoma.
Their results, published earlier this year in Nature, found that high-quality AI can be a major source of support for less experienced clinicians when detecting and treating skin cancer. The study cautioned, however, that faulty AI can mislead even the most tenured researchers in their diagnoses. Ultimately, the researchers see the findings as an important step on the path towards trustworthy and effective human-computer collaboration in their field.