This App Can Detect Signs Of Skin Cancer From Photos Of Your Skin
Using fractal geometry, the app examines the shape and edges of moles for signs of melanoma
Your smartphone isn’t going to replace regular doctor visits anytime in the near future, but if you conduct regular exams of your skin to check for abnormalities or cancerous growths, you’ve got a tool at your disposal in SkinVision, a new app that analyzes photos of your skin, searching for signs of melanoma.
The app’s utility is rooted in fractal geometry, which examines the outer boundaries and shape of a mole for signs that it could be malignant rather than benign. The moles are classified as either low, medium, or high risk based on the algorithm used to analyze photos. Healthy moles are generally symmetrical with smooth edges, while cancerous moles are more likely to be jagged both in general shape and around their edges.
In a test performed by a dermatologist in conjunction with Britain’s Channel 4, Dr. Emma Wedgeworth found the app to be consistently reliable, scanning 195 skin images and correctly diagnosing 73 percent as cancerous and correctly labeling 83 percent as benign.
Obviously, a success rate of 73 percent is hardly foolproof, which is why this is being marketed and recommended as a strictly supplemental tool, rather than a replacement for more traditional medical care. This video, produced by the app makers in 2013, shows how SkinVision, the product of Dutch researchers, works:
It merits noting that in 2015, two similar apps, MelApp and Mole Detective, were forced by the United States Federal Trade Commission to retract claims that their products could detect melanoma. It seems that SkinVision has proven itself more effective over time, but questions abound regarding using your phone as even a supplemental tool in maintaining one’s health.
Simply put by Dr. Ajali Mahto to The Daily Mail, “There is no substitute for a full skin examination by a dermatologist.”