With the sunny days that we have already had this summer, and are yet to come, the prevention and control of skin cancer is again a topical topic in the Netherlands. Just like in Stanford, various Dutch hospitals are working hard to quickly and effectively diagnose skin cancer. At Maastricht UMC+, for example, they recently conducted research, in collaboration with the Catharina Hospital, into a new method to speed up skin cancer control. It concerns a skin scan, the so-called optical coherence tomography (OCT), which is not only faster but also less stressful for the patient.
Detect skin cancer earlier
The development of the Stanford app has a different approach and is therefore primarily intended to offer elderly people, who live in isolation and do not report quickly to the doctor, an opportunity to detect skin cancer in time. A difficult element in the use of the app was the technology, because it was not always easy for the target group to take the right photos with their own hands.
Experts from Skinio, the developers of the mobile application, therefore created a step-by-step guide to taking the photos and helped clinical research coordinators to effectively guide the elderly in using the app on a tablet. The app allows average people and clinicians to take high-quality photos of potentially cancerous lesions without having to go to a clinic. The images are submitted to a dermatologist for assessment via a secure portal.
Algorithm is good tool
“The algorithm for detecting lesions isn’t perfect, but for a triage situation, it’s a good tool,” Kavita Sarin, MD, PhD, an associate professor of dermatology at Stanford Medicine, tells the technology.org website. The study was published in Skin Health and Disease. The app uses machine learning to analyze the photos, scan for skin lesions or abnormalities, and flag those that looked suspicious. But it is explicitly not a diagnostic tool. The app only captures image of skin and analyzes it with AI. The software gives a signal as soon as a spot needs to be examined more closely by a doctor. Of the suspicious spots observed, the app flagged 63% for further investigation. Most of these lesions were ultimately benign.
Three people with skin cancer
Ultimately, 27 elderly people took part in the study. Eleven patients were scheduled for an in-person follow-up visit and four began home treatment. Three people were eventually diagnosed with skin cancer. This could have been missed without the app’s full body images. Some areas of skin cancer had not previously been noticed by the patients.
In the study in Skin Health and Disease we read: ‘In total 88% of patients were satisfied with the Skin Scan program and 77% would probably recommend the program to others. 92% of patients agreed that the pictures from the skin scan were representative of their skin. In the context of the COVID-19 pandemic, 100% of patients found the process safer or comparable to a face-to-face visit. Despite the appreciation for the program, 31% of patients indicated that they would prefer to see the dermatologist after the pandemic.’
–
