|
Briefing
|
- Nail Fungus Diagnosis – South Korean researchers developed deep neural network that diagnoses common nail fungus infection, called onychomycosis
- AI Training – Used Microsoft deep learning algorithm Faster R-CNN to identify and crop images of affected toenails and fingernails, while Microsoft’s ResNet-152 and Oxford University’s VGG-19 model were trained with set of 50,000 images to diagnose nail fungus infections
- AI vs Humans – Algorithms had better results compared to 42 dermatology experts in diagnosis of nail fungus
- Publicly Available – Deep learning approach released in website and as Android mobile app for public to try and experiment with
|
|
Accelerator
|
|
|
Sector
|
Healthcare/Health Sciences, Information Technology
|
|
Source
|
|
|
Original Publication Date
|
February 26, 2018
|
