Google AI application in detecting eye diseases

Recently, Google has announced cooperation with Aravind eye hospital (India) to deploy AI technology in detecting eye disease diseases.

Four years ago, Verily and Google together developed an artificial intelligence system (AI) that could help diagnose diabetic retinopathy (DR), a possible disease. causes permanent eye damage if not treated promptly. In fact, according to the US National Eye Institute, this is the cause of blindness, which is the fastest growing rate for more than 415 million diabetics worldwide. Before this situation, the study by Verily and Google has received great attention from the medical community in general and diabetics in particular, and after four years of hard work, Google finally has It was officially announced that they had successfully implemented and applied AI algorithms in detecting eye pathology at Aravind Eye Hospital in Madurai, India.

Google AI application in detecting eye diseases Picture 1

1. China continues to launch the virtual female radio announcer run by the world's first AI

Previously, Google had detailed details about their partnership with Aravind Hospital in 2017, at the TensorFlow Wired Business Conference and Developers Summit. According to research by Dr. R. Kim, a retinal specialist, Google's AI-based treatment is far more effective than the current efforts of doctors in both respects: Speed ​​and degree exactly.

Accordingly, Google's AI algorithm in the application of treatment of eye diseases has been CE certified, meaning that it meets the European Union standards for medical-related inventions. With the implementation of this new algorithm, doctors at Aravind Hospital will have more time to work more closely with patients in the treatment and control of the situation and development of the disease.

Development of AI-based diabetic retinopathy screening involves managing a huge amount of data, up to more than 128,000 images through which ophthalmologists must analyze and make strikes price of the state of the eyeball. According to Google, these images will be used to train a deep neural network - layers of mathematical functions simulated by biological neurons - to detect symptoms of diabetic retina. The performance of this AI neural network has been tested on two separate clinical trials with a total of about 12,000 images, and the results show that compared to a group of skin and network ophthalmologists. This AI nerve has responded very well and even exceeded the basic performance with an accurate rate of up to 97.5%.

Google AI application in detecting eye diseases Picture 2

1. Samsung introduced Bixby Routines, AI technology that can learn habits and anticipate user needs

In the past year, Google has said it is conducting similar studies in Thailand, and at the Social Benefits Summit in Bangkok the company has also announced a record of cooperation with the hospital. Rajavithi to pilot diabetic retina screening algorithm for patients in the golden pagoda country.

Not only that, Google has also expressed its interest in the field of artificial intelligence applications in health through many million dollar investment deals in health care applications. use AI. Last spring, their Medical Brain Team announced that they had created an AI system that could predict the health situation, the risk of hospitalization, as well as predict the mortality rate at two major hospitals in the United States. States with accuracy up to 90%. In February 2018, Google and Verily scientists created a neural network that accurately deduced basic information about a person's health situation, including their age and blood pressure. It is also possible to determine whether this person is at risk of cardiovascular disease. Most recently, at the end of last year, Google said it had successfully developed an AI system that can detect metastatic breast cancer with almost absolute accuracy, up to 99%.

DeepMind, the division responsible for Google's AI-related studies based in London, United Kingdom, has also participated in a number of health-related AI projects, including an important test that is deployed with the American Veterans Association, involving predicting the health situation of veterans during hospitalization. Previously, DeepMind has also collaborated with the UK National Health Service to develop an algorithm that can detect signs of blindness early. In addition, in an article presented at the Computer and Health conference earlier this year, DeepMind researchers said they have successfully developed an AI system capable of CT scan segmentation. with near performance and level of accuracy equal to the leading specialist doctors.

Google AI application in detecting eye diseases Picture 3

1. AI engineer Facebook talks about deep learning, new programming languages ​​and hardware for artificial intelligence

It can be said that the prospects in the development of artificial intelligence that Google has achieved in recent years have opened a bright future in the thorough and comprehensive application of AI benefits to health. , help improve the quality of health care, human life in a more accurate and efficient way!