1. List of references

| Fact

(500 characters) Conclusion from the fact (200 characters) Author Title Source Year
The first manifestations of brain oncology resemble fatigue, depression and anxiety disorders. And in case of complaints of headache, insomnia and loss of attention, doctors usually advise taking a vacation rather than doing an MRI, especially at a young age. That's why it's so easy
Miss the onset of the disease. Doctors ignore the likelihood of symptoms matching brain tumors, and this leads to the deterioration of patients' condition and misses the optimal moment for treatment. It is necessary to resist this. Mikhailov K.A., Medvedev M.S., Gerasimenko V.A., Tolmachev D.A. ANALYSIS OF THE INCIDENCE OF THE POPULATION OF THE UDMURT REPUBLIC
MALIGNANT BRAIN NEOPLASMS for 2018 https://elibrary.ru/item.asp?id=44379148 2020
Artificial intelligence can be used as a diagnostician and assistant to the attending physician. Having access to databases with thousands and millions of medical records (and other ordered information) using machine learning algorithms, he classifies a specific medical case of a patient, quickly scans the scientific literature on the right topic that has come out over a certain interval, examines the available similar cases and propose a treatment plan. AI will help doctors not to lose sight of the possibility of a patient having a brain tumor, can give recommendations for further examination to the doctor to establish a diagnosis. Bogomolov A.I., Nevezhin V.P., Zhdanov G.A. ARTIFICIAL INTELLIGENCE AND EXPERT SYSTEMS IN MOBILE MEDICINE https://cyberleninka.ru/article/n/iskusstvennyy-intellekt-i-ekspertnye-sistemy-v-mobilnoy-meditsine 2018
In the early stages, brain tumors often cannot be diagnosed. A brain tumor should be suspected if there are the following symptoms:
Progressive focal neurological deficit or general cerebral symptoms
Newly developed seizures
Recent, persistent, unexplained headaches, especially increasing during sleep
Signs of increased intracranial pressure (e.g. optic disc edema, unexplained vomiting)
Pituitary or hypothalamus endocrinopathy To diagnose brain tumors at an early stage, it is necessary to be based on symptoms, as other tests often do not show them. AI analyzing symptoms can help with this. Steven A. Goldman , MD, PhD, Sana Biotechnology Overview of Intracranial Tumors https://www.msdmanuals.com/ru-ru/профессиональный/неврологические-расстройства/внутричерепные-и-спинальные-опухоли/обзор-внутричерепных-опухолей-overview-of-intracranial-tumors?query=внутричерепные опухоли 2021
In Russia, the availability of MRI services is significantly lower than the European average, the availability of diagnostic tests using magnetic resonance imaging is limited by the presence of less than five scanners (4.7 units) per 1 million inhabitants, which is less than a third of the EU average [OECD 2018: 171]. Many patients cannot undergo an MRI examination due to the low availability of machines in the country. Because of this, it is impossible to quickly conduct a study to establish a diagnosis. L. C Panova, A. Yu. Panova Availability of modern medical technologies in Russia and European countries https://cyberleninka.ru/article/n/dostupnost-sovremennyh-meditsinskih-tehnologiy-v-rossii-i-stranah-evropy 2020
Data in many medical organizations are still stored in paper form. This greatly complicates the process of processing them for medical AI training and access to them in general. To work with data, it is necessary to create a portal in which patient data will be entered for further analysis. Portal of the company "Cels” Specifics of medical data for AI training https://celsus.ai/blog/specifika-medicinskih-dannyh-dlya-obucheniya-ii/ 2021
Browser applications are indispensable. They are flexible, versatile, do not require prior preparation of the environment, allow you to save the company's finances, hardware resources, and employee time. The portal should be based on web application technologies. They are convenient, easy to use and work regardless of the power of the PC and are suitable for organizations. AppMaster.io Desktop or web application: pros and cons https://vc.ru/services/297762-desktopnoe-ili-veb-prilozhenie-plyusy-i-minusy 2021
Thus, regions with absolute automation of medical institutions - Kamchatka and Khabarovsk Krai, the Republic of Tatarstan, the city of Moscow, etc. - do not require additional state financial support.
Regions with a share of informatization from 80 to 99% - Vladimir, Lipetsk, Novosibirsk regions, St. Petersburg, etc. Many regions have reached a high level of informatization and the implementation of the portal will not be very difficult. Fedyaeva Olesya Alexandrovna Informatization of the health care system in Russia: problems and proposals for their solution https://cyberleninka.ru/article/n/informatizatsiya-sistemy-zdravoohraneniya-v-rossii-problemy-i-predlozheniya-po-ih-resheniyu 2019
1. Information on the fact of a citizen's application for medical care, his state of health and diagnosis, other information obtained during his medical examination and treatment, constitutes a medical secret.2. It is not allowed to disclose information constituting a medical secret, including after the death of a person, by persons to whom they became known during training, performance of labor, official, official and other duties, except in cases established by parts 3 and 4 of this article. Anamnesis, life indicators, results of tests and images are a medical secret, which is forbidden to disclose, and therefore according to Part 3 of Art. 13 FZ 323, consent to the disclosure of doctor's secrets is required. Russian State Duma Part 1,2,3 st. 13 of the Federal Law of November 21, 2011. N 323-FZ "On the basics of protection of citizens' health in the Russian Federation” https://student2.consultant.ru/cgi/online.cgi?req=doc&base=LAW&n=412859&dst=100101%2C1&date=15.06.2022 Latest edition from 2022.
Processing of personal data - any action (operation) or set of actions (operations) performed with or without the use of automation tools with personal data, including collection, recording, systematization, accumulation, storage, clarification (update, change), extraction, use, transfer (distribution, provision, access), depersonalization, blocking, deletion, destruction of personal data; The mandatory depersonalization procedure for the use of data is the processing of personal data, which means that consent to the processing of personal data is required for the use of data Russian State Duma Paragraph 3 of Part 1 of Article 3 of the Federal Law of July 27, 2006. N 152-FZ "On Personal Data” https://student2.consultant.ru/cgi/online.cgi?req=doc&base=LAW&n=389193&dst=100239&date=15.06.2022 Latest edition of 2021
Existing tested AI services use PET-CT, ultrasound, MMG, CT, MRI, ECG, FLG, WG To use AI services, you need access to equipment, surveys. There are no systems that worked on analyzes and symptoms Scientific and Practical Clinical Center for Diagnostics and Telemedicine Technologies of the Moscow Department of Health Catalogue of AI services https://mosmed.ai/service_catalog/ 2023
Life expectancy after surgical treatment for brain cancer directly depends on the stage at which the operation was performed. The best predictions regarding survival and life expectancy are observed in stage 1 of cancer, provided that the disease has not relapsed in the future. If surgical care was provided at an early stage, the life expectancy after surgery can be up to 5 years. The sooner the tumor is detected, the more effective the treatment will be, and the longer the patient's life expectancy. Alexandra Tomberg How quickly a brain tumor develops. Life expectancy https://clinics.direct/ru/articles/how-fast-a-brain-tumor-develops/#:~:text=В среднем эта цифра составляет,и стоимость лечения рака мозга. 2019
68% of medical professionals believe that the waiting period for complex and highly informative diagnostic examinations (magnetic resonance tomography (hereinafter referred to as MRI), computed tomography (hereinafter referred to as CT), angiography) significantly exceeds the established PGG 30 days MRI and CT diagnostics take a very long time due to the long waiting queue, which means that the diagnosis of MRI and CT will confirm the diagnosis too late. I.M. Sheiman, S.V. Shishkin (heads of the author's team),
V.O. Kusakina, V.O. Moiseeva, S.V. Sazhina, E.A. Tarasenko, A.E. Chirikova WAITING TIMES
MEDICAL
AID:
FOREIGN EXPERIENCE
AND RUSSIAN
Practice https://www.hse.ru/data/2019/04/12/1178006146/10 Сроки_ожидания_медпомощи.pdf 2019J

To solve the task, it is necessary to create a web portal that will analyze the history and tests, not only MRI, which will be able to detect the tumor at an early stage, paying attention to everyday symptoms. All data must be stored on certified servers, all documents for data must be received.

2. Main information about project

Name Expert system for the determination of internal tumors
General concept To solve the problem, a web portal has been created, which can be used as a separate system. To work with the portal, each patient will give explicit consent to the processing of PD, consent to the disclosure of medical secrets in an impersonal format to improve the diagnosis and operation of ES.
Target audience The first subsystem is small clinics in small towns, rural paramedic stations that do not have CT and MRI, reception rooms of hospitals of any size, polyclinics
The second subsystem is oncology centers, large hospitals with the equipment necessary for high-flow diagnostics
How it works? The system is divided into two subsystems - risk assessment, diagnostics. The first helps to calculate the risk of OGM at the initial appointment with a doctor in small hospitals without the necessary equipment, based on the symptoms. The second helps to accurately diagnose OGM on high-precision equipment in large clinics.
Benefits of the idea Patients will be able to avoid deposits by identifying the risk of developing OGM at an early stage, even in remote hospitals.

3. Implementation Steps

Steps Time to perform Expected result
Assembly of the development team, creation of terms of reference, documentation 1 month Команда разработчиков собрана, понимает суть проекта, необходимые технологии, готова к созданию портала.
Creation of the MVP portal, initial testing 6-12 months The MVP of the portal has been developed, it can be used for demonstration purposes, all primary tests have been passed (you can add/edit/delete accounts, patient cards)
Finalization of the portal to a fully functional version 3-6 months Final version of the portal, ready for implementation in medical institutions
Applying for a grant from Yandex Cloud 14 days Application documents will be ready, application submitted
Receiving a grant from Yandex Cloud, creating a project in the cloud 21 days (7 to receive, 14 to create) Object Storage Yandex Cloud database created, paid for grant funds
Integration of images and anamnesies into the database 14 days Dataset for AI has increased, AI accuracy has increased
Preparation of consents to PD, disclosure of medical secrets 7-14 working days Consents for the user have been developed
Purchase of hosting and domain, portal placement 7 days The portal is available to users, published on the Internet
Creation of a training course for employees of medical organizations, users 1 month Designed, written course, ready for publication
Preparation for testing of the portal at the pilot site 14 days An organization account, staff profiles have been created on the portal, a data transfer system has been established, the necessary personnel have been trained to work with the portal
Testing of the portal on the pilot site 1 month Deficiencies in everyday use have been identified, feedback from staff and users has been received, the first data has been obtained, the first users have been registered
Development, updating, promotion of the portal on state websites Until the end of the implementation period The portal will be improved, developers will fix vulnerabilities and prevent data leaks.

4. Budget

Ресурс Resource cost/ resource quantity Source
Object Storage Yandex Cloud service for storing anonymized images, research results, corresponding to FZ 152. 15,903.19 rubles per month, for configuration: 10 TB cold storage, with quotas: 7,200,000 GET operations, 3,600,000 POST operations, 100 GB of outgoing traffic. Grant from the Yandex Cloud platform - Yandex Cloud Boost, Start program.
Payment for the hosting platform from REG.RU 1,619 rubles per month for VIP-1 server (suitable for loaded sites) Grant
Payment for a domain name from the domain registrar From 199 rubles per year Grant
Development of the portal 300,000 - 6,000,000 rubles Grant
Legal services for drawing up consent to the processing of PD and disclosure of medical secrets 6,000 - 15,000 rubles Grant
Creation of a training course for medical workers and users. 10,000 - 15,000 rubles Grant
Payment of the development staff to support the database 100,000 - 160,000 rubles per person Grant
Time to develop the portal 6 months - 1 year Developers
Time to write consents 7 - 14 working days Lawyer
Time to develop a course on use 1 month Course developers
Time to test the system 1 month Hospital workers
Time to receive a grant from Yandex Cloud Paperwork 14-30 working days receiving a grant 7 days Representatives of Yandex Cloud, representatives of the platform
Time to embed snapshots, survey results into the cloud 6-12 months Representatives of medical institutions, developers

5. Presentation

https://youtu.be/P_oUiyVkToU

(in Russian)