Facial Recognition

Facial recognition is a bio-metric recognition technology, which has the advantages of non-intrusiveness, non-contact, friendliness, and convenience. This technology uses computer vision algorithms to analyze facial photos and recognizes genetic diseases with facial features to diagnose genetic diseases through facial recognition.

Early diagnosis of genetic diseases is very important for treatment, but accurate diagnosis is often a long and expensive process. The correct diagnosis and early treatment of genetic syndromes caused by gene mutations often rely on the experience of doctors. However, even experienced clinicians can hardly diagnose genetic syndromes of different races. Our artificial intelligence system uses face recognition technology combined with neural networks to intelligently recognize facial features that can help doctors assist in diagnosis.

Model Components

DeepGestalt deep learning algorithm
Our facial recognition model is based on the DeepGestalt deep learning algorithm, which combines computer vision and deep learning algorithms.

It is a new type of facial analysis framework that can distinguish the contours of a variety of genetic syndromes, such as congenital thymic dysplasia, total forebrain malformation, Rubinstein-Tabby syndrome, cholesterol alcohol syndrome, etc.

As shown in the figures below, the DeepGestalt deep learning algorithm lists the syndromes that each facial image may represent and sorts them according to the probability of different syndromes. Experimental results show that the DeepGestalt deep learning algorithm has a 90% probability of successfully listing the correct disease name in the first 10 answers.

2 Data collection
To train such an algorithm requires a considerable data set, we collect data through publications and public data sets, as well as through our applications. On the one hand, our system helps doctors diagnose, on the other hand, it collects the facial photos of patients uploaded by the doctors and enlarges the database.

Analysis process of DeepGestalt algorithm

Figure a Analysis process of DeepGestalt algorithm

Phenotype extraction process of DeepGestalt algorithm

Figure b Phenotype extraction process of DeepGestalt algorithm

Advantages

1 Accurate
The world's leading recognition accuracy has been tested by various industries. In clinical applications, the artificial intelligence algorithm Deepgestalt has an accuracy rate of 91% in identifying the correct syndrome on 502 different images.

2 Fast
Use cloud services or offline capabilities, no need to wait, upload facial photos anytime, anywhere, and get diagnosis results instantly.

3 Reliable
Our facial recognition combines computer vision and deep learning algorithms. This is a new type of facial analysis framework that is better than clinical experts in independent experiments. Furthermore, supporting high concurrent hosting and high availability, our servers can provide you with a full range of service guarantees.

MedAI has formed a team of experts excellent in imaging science and clinical domain knowledge, providing AI-driven solutions for facial recognition according to your detailed requirements.

Online Inquiry

Please submit a detailed description of your project. We will provide you with a customized project plan to meet your research requests. You can also send emails directly to inquiry@protheragen.com for inquiries.

Verification code

Copyright © 2021 Protheragen Inc. All Rights Reserved.