[an error occurred while processing the directive]
RSS
Логотип
Баннер в шапке 1
Баннер в шапке 2
Project

Care Mentor AI using the supercomputer of Skoltech will create service for determination of extent of defeat of COVID-19

Customers: Care Mentor AI (Kerementoreyay)

Product: Zhores Superkompyyuter

Project date: 2020/02  - 2020/08

2020: Use of the Zhores supercomputer for determination of extent of defeat of COVID-19

Within the program for fight against COVID-19 of the Center of Skoltech for scientific and engineering computing technologies for tasks with data bulks (CDISE) date scientists of Care Mentor AI company could use the Zhores supercomputer for increase in accuracy of determination of pathologies. On September 2, 2020 the company reported about it Care Mentor AI.

Use of the Zhores supercomputer for determination of extent of defeat of COVID-19
.
File:Aquote1.png
Thanks to the Zhores supercomputer we could make for a short time a set of experiments and receive a neuronet with high quality for our service of determination of severity COVID-19 on complete 3D KT-researches of bodies of a thorax. I am sure that our further cooperation will lead to improvement of services of computer vision Care Mentor AI, thereby reducing load of radiologists and allowing to carry out diagnostics of patients quicker, – Pavel Roytberg, the co-founder of Care Mentor AI companythe Russian developer of services of computer vision for the analysis of researches in the field of radiodiagnosis on the basis of neural network algorithms notes.
File:Aquote2.png

The product Care Mentor AI is the system of artificial intelligence for the analysis and interpretation of results of beam methods of a research. The neural network technology analyzes results of radiodiagnosis and with a high speed and accuracy defines presence or absence of pathologies at patients, including malignant new growths, tuberculosis, pneumonia and so forth. Also a system analyzes beam researches on screening of pathologies and determines percent of defeat and severity of COVID-19 by the analysis of KT-researches. The computer tomography allows to detect more precisely the pathological signs inherent in virus damage of lungs, first of all, interstitial infiltration of pulmonary fabric that it is quite difficult to make on the roentgenogram. Besides, only KT allows to define objectively localization and prevalence of pathological process and transformation of one stage of a course of a disease in another that is extremely important for assessment of weight and dynamics of pathological process.

Zhores is the petaFLOPS energy efficient supercomputer which is specially intended for solving of tasks of the machine learning and modeling based on data. This computing system is aimed to help scientists of Skoltech both its academic and industrial partners to perform breaks in medicine and other areas.

File:Aquote1.png
Developing the Zhores supercomputer, we put in it architecture which meets job requirements with data bulks, especially in the field of work with data of biomedical researches. It allowed us to train neuronets of Care Mentor AI company which are used for the analysis of pathologies on radiodiagnosis researches. Strengthening opportunities and developments of our partner, we together will be able to bring medicine of Russia to other technology level. Biomedical data have huge dimension therefore for work with them computing powers with a possibility of a parallelization of all processes are required. Thus, our developments outstripped the time thanks to architecture of the Zhores supercomputer a little, – Maxim Fedorov, the vice president of Skoltech in the field of artificial intelligence and mathematical modeling comments.
File:Aquote2.png

Specialists of the project participant Skolkovo of Care Mentor AI report that the company already taught a neuronet "see" cancer centers on KT-researches with an accuracy of 95%, calculate percent of damage of lungs at COVID-19 infection with an accuracy of 86% and also calculate slope angle of foot and set existence or absence of flat-footedness with an accuracy of 99%. Moreover, the Care Mentor AI service is able to prioritize and mark with an accuracy of 93% pathologies on radiological researches, helping the doctor to cope quicker with the volume of work and to manage to conduct the bigger number of diagnostic testings.

A system on radiographic screening of pathologies of bodies of a thorax passes pilot test in JSC Meditsina – clinic of the academician Roytberg. Also the product on radiographic screening of pathologies of bodies of a thorax passes test within the Moscow experiment on use of the innovative technologies in the field of computer vision for the analysis of medical images and further application in a health care system of the city of Moscow. During this experiment service is integrated in more than 45 medical institutions of Moscow.

The Care Mentor AI system according to the analysis of KT of researches works for determination of percent of defeat and severity of COVID-19 in the reference center of the Ivanovo region where all KT-researches with suspicions of a coronavirus come from all hospitals of the Ivanovo region.