A deep learning algorithm to detect the presence of basal cell carcinoma on Mohs micrographic surgery frozen sections

Abstract

To the Editor: Nonmelanoma skin cancer represents the most common cancer in the United States with more than 5.4 million cases annually, of which 8 in 10 are basal cell carcinomas (BCCs).1 Mohs micrographic surgery (MMS) treats BCCs by providing margin control. The process of accurate and complete interpretation of frozen section slides would benefit from the presence of a second reader. We seek to apply deep learning to MMS by developing an algorithm to aid the surgeon in detecting BCC on frozen sections.