Can an app warn you that you might have skin cancer? Digital health startup SkinVision is using smartphone tech and vision algorithms to allow consumers to track changes to their moles. Users of its iOS and Android apps, which have been downloaded some 200,000 times since launch back in 2011, take a photo of a mole and the app then performs a visual analysis, with the aim of identifying suspicious growths such as melanoma.
“Skin cancer grows chaotically and potential suspicious moles are identified based on signs of non-natural growth,” explains CEO Dick Uyttewaal. “The algorithm within the online assessment reviews signs of non-natural growth of skin lesions and is based on an established mathematical methodology in biology called fractal geometry.
“The algorithm currently looks at seven different criteria and will be further improved based on the continuous growth in our database (currently in excess of 1 million pictures).”
Moles are rated using a simple traffic light system (using a red, orange or green risk rating). The app lets users store photos in multiple folders so they can track different moles over time.
“A changing mole (color, size, symmetry etc.) is a clear sign that something is wrong and that the person should visit a doctor immediately,” he adds.
How accurate is the SkinVision tech at identifying melanoma? Uyttewaal says it’s as good as the “average eye of a dermatologist”, noting also that it’s the only such skin cancer detector app to have obtained CE certification in Europe.
Read more at YAHOO