Skin Image Search™ to Help Identify Any Skin Concern with a smartphone
San Francisco, CA – April 24th, 2018 – First Derm® launches its FREE Artificial Intelligence (AI) API to help assess any skin concern within a second. Their Skin Image Search™ uses the API and is in beta and will be improved on a monthly basis, by continuously training a deep Convolutional Neural Network (CNN) with skin disease images.
Research backed[1] First Derm® started out as an anonymous, on-demand-online dermatology app in 2014 and has since collected hundreds of thousands anonymous, amateur skin disease images. The images have been used to improve the service and for scientific research; where using AI is the new frontier. First Derm’s ongoing scientific research has preliminary shown similar results to other AI skin disease studies[2],[3] that have used trained CNNs.
“We started out as an asynchronous teledermatology service – as CNNs have evolved we have been well positioned to use our unique data set. Thanks to an awesome team, we now have a great tool that can be used as a symptom checker for any skin concerns that we have named Skin Image Search™. It is still in beta and gets better by every month”, says Dr. Börve, CEO and founder of First Derm®.
Try our Skin Image Search™ here
How We Are Different to Other Dermatology AIs
First Derm’s AI API, powers an online Skin Image Search™ tool available online in any web browser. The Skin Image Search™ has been trained on images of any skin condition; ranging from skin rashes, hair loss, nail issues, skin cancer lesions, and genital concerns – including visual STDs. The database contains one of the largest datasets of unique amateur smartphone images, taken by the everyday person. The unique, real-life dataset can help match for similar images. By comparison, other AIs have been trained on perfect clinical medical images, focusing mostly on skin cancer and does not represent the image quality that a layman would capture.
The AI API is user friendly and can be easily integrated into any website, app, or electronic healthcare record (EHR). First Derm® has already partnered with an online primary care telemedicine service that uses the API to help guide their family doctors to make better skin disease assessments for their online video patients.
First Derm’s Skin Image Search™ is in beta and will progressively increase sensitivity and specificity as more images train the CNN. Currently 80% of the volume of today’s asynchronous First Derm® customers, covers thirty-three of the most common skin concerns. The Skin Image Search™ accuracy rate for assessments is at approximately 80%, when five differential skin diseases are included in the search answer. The Skin Image Search™ does at this stage not replace a dermatologist’s diagnostic skills, but it is a good practical aid that can help for guidance on any skin concern – it is more powerful than just an internet search comparing images. Skin images on the internet are labeled incorrectly in about 50% of the cases.
“Already today we are better than general doctors. In nine months time, with our continued research, we anticipate the AI to be comparable to the diagnostic accuracy of a dermatologist on any skin concern. Our preliminary Skin Image Search™ test-results have already caused some diagnostic debate within our group of dermatologists”, says Dr Börve.
Try our Skin Image Search™ here
###
[1] FirstDerm Resarch https://firstderm.com/teledermatology-background-research/
[2] Esteva A, Kuprel B et al Dermatologist-level classification of skin cancer with deep neural networks. Nature 017/02/02/
[3] Seung Seog Han et al Deep neural networks show an equivalent and often superior performance to dermatologists in onychomycosis diagnosis: Automatic construction of onychomycosis datasets by region-based convolutional deep neural network PLOS ONE | https://doi.org/10.1371/journal.pone.0191493 January 19, 2018
The Specialist doctor from the University Hospital in Gothenburg, alumnus UC Berkeley. My doctoral dissertation is about Digital Health and I have published 5 scientific articles in teledermatology and artificial intelligence and others.
