Are Apps Capable of Accurately Diagnosing Skin Cancer?
Posted: Tuesday, September 18, 2018
The haste to develop successful apps for skin cancer diagnosis may risk public safety, based on a study reported at the British Association of Dermatologists’ Annual Meeting in Edinburgh (Abstract BT08) and published in the British Journal of Dermatology. Delaying skin cancer diagnosis causes significant declines in survival rates, and although some apps have demonstrated high success rates, the study authors highlighted three major failings: a lack of published trials showing safety and efficacy, a lack of input from specialists during app development, and flaws in photo analysis.
“Future technology will play a huge part in skin cancer diagnosis,” Maria Charalambides, PhD, of the University of Birmingham’s College of Medical and Dental Sciences, stated. “However, until adequate validation and regulation of apps is achieved, members of the public should be cautious when using such apps, as they come with risk.”
The authors explored the benefits and limitations of a number of apps on the market, including teledermatology, photo storage, and risk calculation. Some apps showed high success rates. For example, teledermatology and fractal theory analysis correctly identified 88% and 73% of people with skin cancer and 97% and 83% of those with benign lesions, respectively.
However, color and pattern recognition software apps seem to be less accurate in recognizing scaly, crusted, or ulcerated areas or amelanotic melanomas. Apps that compare images on a monthly basis or advise users to seek dermatologist review based on risk calculation appear to be unable to assess finer details that may be identified using a dermatoscope or in person. These issues may increase false-negative results and often delay treatment.
“Apps can currently supplement but not substitute standard medical care,” the investigators concluded.
