Press release – Artificial Intelligence (AI) Dermatology Algorithm API – Version 1.3
Skin Image Search™ can identify 33 skin diseases with 80% accuracy
SAN FRANCISCO, CA (PRWEB) June 13, 2019 – First Derm® releases version 1.3 of its Artificial Intelligence (AI) API to help assess skin concerns within a second, it can be used for FREE by the consumer as a Skin Image Search™ tool. 100 000 annotated images on 33 different skin diseases and in total over 300 000 images on 248 different skin diseases have been used to train a deep Convolutional Neural Network (CNN).
Dr Alexander Börve was recently interviewed on 90.9 WBUR-FM Boston’s NPR news station about how he started out with telemedicine research and set up First Derm® a global on-demand-online dermatology service. He also spoke about the future of teledermatology, which he is predicting will all be done by computers (AI), and will be faster, more accurate and cheaper than a human dermatologist.
“I have been blown away by how good CNNs can pick up skin diseases. I was predicting that it would take 24 to 36 months to be better than a dermatologist, but if we continue at this pace we will be better within the next 12 months. Thanks to an awesome team, we are making rapid progress”, says Dr. Börve, CEO and founder of First Derm®.
248 skin diseases and 300 000 images
First Derm® have done Machine Learning (ML) on 100 000 smartphone annotated images and over 300 000 smartphone skin diseases, ranging from inflammatory diseases such as acne and psoriasis, to skin cancer, like malignant melanoma and nevus, to STIs such as Molluscum Contagiosum and genital herpes. The database has one of the largest datasets of unique amateur smartphone images and First Derm® is collecting another 15 000 images per month. By comparison, other AIs algorithms have been trained on perfect clinical medical images, focusing on skin cancer, which does not represent the image quality that a layman would capture.
China, Russia, Vietnam, telemedicine and chatbot integration
First Derm’s just signed an agreement with Tencent to get a verified Chinese mainland WeChat account and will release their AI as a bot under idoc24在线人工智能皮肤问题咨询 . Other partnerships will be announced soon.
Continued scientific research needed
Our AI algorithm accuracy has increased with about 20% since version 1.0 and one of the 33 skin diseases will be in the top 3 with an about 80% probability. Further research is still needed and scientific clinical studies are needed, before we can confirm our accuracy with certainty. So for now we market the service as an image search on skin diseases – comparing uploaded photos to our database that we have trained our AI on. First Derm® is in the process to seek FDA approval for its dermatology algorithm.
About First Derm®
First Derm® leads the mobile health revolution by empowering users to submit skin aliments anonymously to board-certified dermatologists with photos and a symptom description, from anywhere in the world, in any internet connected device. First Derm® is backed by scientific research as a triage service, yielding an information-based response and suggested course of action. Since the service can be accessed any time in any internet connected device, users receive a reliable review of their ailment directly from a board-certified dermatologist, even on holidays and weekends. The app and parent company, iDoc24 Inc. are headquartered in San Francisco. First Derm® is available for free download on both iOS and Android, as well via our online widget. For more information, please visit http://www.FirstDerm.com . Try our AI here: Skin Image Search™ – Artificial Intelligence (AI) – Dermatology – FREE
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.