Sciberia System for recognizing pathologies in medical images using AI

The main articles are:

• Machine Learning
• Artificial Intelligence in Medicine
• Image Transfer and Archiving Systems (PACS)

Sciberia is a diagnostic system that allows you to automate the processes of recognizing various pathologies in medical images using AI.

The company 'Cyberia' has developed a method for analyzing medical images computed tomography based on technology machine learning and computer vision. The development of'Cyberia' allowed to automate the processes of recognition of various pathologies in medical images, which in turn helps to accurately and faster diagnose the patient. Today, the accuracy of diagnostics of the system is 98.5%. The speed of processing images is 15 seconds.

2022: Mishustin instructed the Ministry of Health to introduce into clinical practice the Russian IT system for analyzing medical images Sciberia

At the end of August 2022, Prime Minister Mikhail Mishustin will introduce a Russian IT system into clinical practice to analyze medical images of Sciberia. This was reported on the website of the Government of the Russian Federation.

According to Mishustin's instructions, the Ministry of Health of the Russian Federation until April 1, 2023 will have to "analyze the feasibility of introducing the Sciberia complex for processing and analyzing medical images and the Onkopoisk online medical examination project and the need to register them as medical devices in the manner established by the rules for registering medical devices."

Mikhail Mishustin

The Yakut company Sciberia specializes in the development of software for the analysis of medical images of CT, magnetoresonance tomography and radiographs using machine learning technology.

In 2022, the company announced the development of comprehensive software (software) for medical institutions. It includes tools for analyzing and evaluating CT images, storing and processing medical images, cloud services for working with software modules.

The Onkopoisk screening project, which appears in the order of Mikhail Mishustin, is also a Yakut initiative. It is being implemented within the framework of the regional project "Fight against oncological diseases" of the national project "Healthcare" on the basis of the Yakutsk Oncology Center.

The developments of two Yakut companies in the field of medicine were presented to Mikhail Mishustin during a visit to the IT park in mid-August 2022. The prime minister noted the importance of these developments in terms of the implementation of the import substitution program.

The COVID-19 coronavirus pandemic has seriously spurred the development of technologies, including the development of artificial intelligence technologies. I know that you have serious technologies for recognizing covid and lung diseases, MRI and brain diseases. I definitely ask the Ministry of Health of the Russian Federation, in particular Mikhail Albertovich Murashko, to look at all the developments, because the generalization and standardization of these things is very important. Well, God forbid that people who live in remote areas can get diagnosed very quickly and efficiently. Today, correct diagnostics removes a huge number of questions, including the navigator for doctors who can provide assistance in remote places, "Mishustin said then.[1]

2021

Implementation plan in Europe, North America and Taiwan

The medical diagnostics module from Yakut developers will be implemented in Europe, North America and Taiwan. This was announced on August 19, 2021 by the Press Center of the Republic of Sakha.

As of August 2021, the Syberia diagnostic system in pilot mode has already been tested in 16 hospitals in Russia, 3 of which continued its use. By mid-August 2021, Sciberia had raised about $300 thousand in investments in its project.

In April 2021, our team passed the international program on innovative entrepreneurship YEDI International Program, and now in partnership with the public organization from Canada YEDI (York Entrepreneurship Development Institute) "Cyberia" intends to enter the Canadian market and pass product certification to start sales in North America. And in June 2021, we were invited to participate in the Taiwan Program for Technology Startups to Enter the Asian Market A2T (Access to Taiwan). The program provides for the passage of several accelerators and provides networking with large corporations, investors, entrepreneurs, industry experts, which allows you to implement a test entry into the market of Taiwan and in the future of other Asian countries. We will be given the opportunity to receive funding and state support for Taiwan, soft-landing services and co-working space, a database of the necessary contacts will be opened, and the adaptation and development processes are accompanied by experienced mentors and mentors, "said Afanasy Argunov, founder of the Cyberia company.

Kiki Lin, a representative of the Access to Taiwan (A2T) program, who is responsible for establishing ties between customers and startups in A2T, noted that the program experts will help the Yakut startup develop in the Taiwan market.

Cyberia has developed a method for analyzing medical images based computed tomography on the technology machine learning and. computer vision The development of 'Cyberia' allowed the automation of the processes of recognition of various pathologies in medical images, which in turn provides an opportunity for a more accurate and rapid diagnosis of the patient. The diagnostic accuracy is 95%. Total diagnostic time: 45 seconds. The system has already processed more than 10 thousand. CT studies.

In 2021, the company applied its developments to create a module for automated detection of signs of acute cerebrovascular accident, such as ischemic and hemorrhagic stroke. The stroke diagnostic module from Cyberia will significantly reduce the time to detect the disease and timely prescribe thrombolytic therapy.

Sciberia Diagnostic System Composition

As of August 2021, the Sciberia program complex consists of the following components:

• Sciberia Viewer - a tool for a doctor, designed to view medical images, allows you to view images, conduct initial assessment and measurements;
• PACS software Sciberia - for storages and transmission of medical images, allows you to download, store and transfer medical images both within one medical organization and between different medical organizations;
• Sciberia Lungs - a software module for processing CT images of the chest organs and quickly forming their characteristics for the diagnosis of COVID-19 and lung cancer;
• Sciberia Head - a software module for processing CT images of the brain and quickly forming their characteristics for the diagnosis of strokes;
• Sciberia WS - cloud services for working with software modules Sciberia Lungs and Sciberia Head, there is also functionality of a personal account for a doctor and employees of a medical organization.

Creating a Module to Detect Stroke Signs

Sciberia, a graduate of the B8 accelerator, a resident of the Yakutia Technopark, Yakutsk IT Park and Skolkovo, will release a diagnostic module for detecting signs of acute cerebrovascular accident - stroke. The Press Center of Yakutia in Moscow announced this on March 31, 2021. The module works using machine learning technology: the system independently analyzes the CT image, which allows the doctor to automate and significantly speed up the process of detecting and localizing signs of ischemic and hemorrhagic stroke, that is, improve the quality of medical care provided to patients. For March 2021, the team has begun final development, the module is planned to be released approximately by the fall of 2021. The project is being implemented at the initiative of Sciberia at the expense of the company's own funds.

Ischemic stroke is one of the common causes of mortality and disability. The diagnostic module will allow the specialist to quickly scan the patient and prescribe thrombolytic therapy in a timely manner, if necessary, - said Afanasy Argunov, co-founder and technical director of Sciberia.

The diagnostic module uses a multimodal viewer - a cross-platform (Windows, Linux, MacOS) tool designed to view and analyze medical images - with its help the device produces an examination result. A multimodal viewer (image viewer) developed by Sciberia, it is an important component of diagnostic modules for determining various diseases. The program can work as a separate tool at the doctor's workplace, as well as as as part of a diagnostic complex with other software products - a PACS server, a cloud web service and with different diagnostic modules based on artificial intelligence. The multimodal viewer is being tested in the work of the COVID-19 module for the diagnosis of coronavirus pneumonia in medical institutions in Yakutia. As of March 2021, the Sciberia COVID-19 module has processed more than 10 thousand chest CT studies and has a sensitivity of 95%.

Our team initially dealt with a system for processing chest computed tomography studies to detect rounded lesions. With the onset of the pandemic, we found that our algorithms are suitable for screening for viral pneumonia caused by coronavirus infection. We quickly developed and launched a COVID-19 screening module within our system. As the COVID-19 module is introduced, we continue to develop modules of the IRIS complex, including a multimodal medical image viewer, and have begun the final development of a module for diagnosing stroke, "Afanasy Argunov shared his creation story.

2020: Application

In 2020, the development was used for applied tasks to identify nodular lung formations and to identify signs of viral pneumonia COVID-19. During the pandemic, the Syberia coronavirus pneumonia diagnosis system was tested in 16 hospitals in Russia, three of which continued to be used.

See also

• Brain stroke
• Lung cancer
• Coronavirus COVID-19