Xidian University has developed an "electronic doctor" to screen congenital cataracts
Release date: 2017-05-05
It's like cloning an experienced ophthalmologist to complete an automated diagnosis through a cloud service. Recently, Professor Liu Xiyang from the Software College of Xidian University and the team of Dr. Lin Haotian from Zhongshan Ophthalmic Center of Sun Yat-sen University have developed a prototype CC-Cruiser for congenital cataract screening. The research results are on Nature. Subtitle "Nature Biomedical Engineering".
Liu Xiyang said that for rare diseases that are difficult to obtain high-quality data, how to make artificial intelligence into the "right arm" of the doctor in the process of screening and assisting diagnosis? Therefore, Liu Xiyang team and Lin Haotian team began a collaborative research on congenital cataract machine diagnosis in March 2015. By refining doctors' diagnostic experience, the researchers developed a diagnostic system CC-Cruiser and assisted doctors in the automated diagnosis of congenital cataracts in the form of cloud services.
“Working with us is the top ophthalmology center in China. The doctors have rich experience in diagnosis and treatment, bringing together the largest number of congenital cataract patients in the world.†Liu Xiyang said, “Our research is to treat this kind of quality doctor. Experience is condensed into the machine, allowing the computer to conduct deep learning through the form of training samples, using the system we created to give more patients the opportunity to receive high-level medical diagnosis and treatment."
In the study, the CC-Cruiser collection used partial images from the routine examination of the Chinese Ministry of Health's Children's Cataract Program. All images were independently classified and described by two experienced ophthalmologists. The third ophthalmologist disagreed. supply enquiry. After completing the training, CC-Cruiser can help screen patients with congenital cataracts and conduct risk assessments on patients.
Liu Xiyang told reporters that the CC-Cruiser performance was tested five times and the results were excellent. CC-Cruiser made a correct diagnosis in a comparison of 50 images with human ophthalmologists. CC-Cruiser's performance is called a "qualified ophthalmologist." He said that if you clone a large number of artificial intelligence ophthalmologists who meet or even exceed the doctor's average, doctors only need to confirm the results and diagnose them again when they encounter a difficult condition. Check efficiency and quality will be greatly improved.
During the research process, Liu Xiyang team developed an information management system for eye diseases to help automatically collect patient medical records and image data. Through the use of information systems, doctors will continue to collect new medical records and new disease data, which are all follow-up. Research provides a source of data. In order to benefit patients in remote areas, the research team also developed a set of auxiliary diagnostic software to move the trained model to the “cloud†for medical service management of congenital cataract. On this system, ordinary users can also upload their own eye images, and the website will immediately give a diagnosis. To prevent misdiagnosis and missed diagnosis, a professional doctor will check all uploaded cases every week.
Source: Xi'an Evening News
Bluetooth TWS hearing earphones for impairment
Shenzhen Sunshine Technology Co.,Ltd , https://www.shenzhenyatwin.com