In this diagnostic/prognostic study, Han and colleagues evaluated a deep learning algorithm designed to identify facial lesions and predict risk of skin cancer. A total of 924 538 training image crops, including of various benign lesions, were generated with the help of a region-based convolutional neural network. Results suggest that the algorithm can localize and diagnose skin cancer without preselection of suspect lesions by dermatologists. Tschandl provides an Editorial.

Editorial

Taube and colleagues performed a nonrandomized controlled trial to investigate the association of bariatric surgery with skin cancer incidence. The study included 2007 patients with obesity who underwent bariatric surgery and 2040 contemporaneously matched controls who received conventional obesity treatment. Findings showed that bariatric surgery is associated with a reduction in the incidence of skin cancer, including melanoma, and that there may be an association between obesity and this cancer form.

在这项诊断/预后研究中，Han及其同事评估了一种深度学习算法，旨在识别面部病变并预测皮肤癌的风险。借助基于区域的卷积神经网络，总共生成了924 538个训练图像作物，包括各种良性病变。结果表明，该算法无需皮肤科医生预先选择可疑病变即可定位和诊断皮肤癌。Tschandl提供了社论。

社论

Taube及其同事进行了一项非随机对照试验，以研究减肥手术与皮肤癌发病率的关系。该研究包括2007年接受减肥手术的肥胖患者和2040名同时接受常规肥胖治疗的对照患者。研究结果表明，减肥手术与减少皮肤癌（包括黑色素瘤）的发生率有关，并且肥胖与这种癌症形式之间可能存在关联。