RSS
Логотип
Баннер в шапке 1
Баннер в шапке 2

Third opinion

Company

width=200px

Owners:
Medsi - 12,5%
Sberbank

Owners

"Third Opinion" - an intellectual system built as a neural network is able to determine the type of pathology from the relevant medical image online with an accuracy of 75-95%. For neural network training, medical assays must be pre-collected, digitized, and marked up as images. Currently, the network is trained in the following types of images: blood cell images, fundus images, lung radiographs, bladder ultrasound. The plans for 2018-2019 include marking and subsequent recognition of pathologies by CT and MRI.

History

2020

Medsi bought 12.5% of Third Opinion

On June 19, 2020, it became known that the network of private medical clinics Medsi"" was bought by 12.5% of Platform Third Opinion LLC (PTM), which develops artificial intelligence algorithms for use in the medical field. The value of the transaction was not disclosed.

The Third Opinion is developing a SaaS platform for recognizing pathological changes in medical images, including bone marrow smears, fundus images, chest X-rays and computed tomography (CT), mammograms, etc. Analyzing medical images by a person can take from a few minutes to an hour. The Third Opinion software platform recognizes the problem in a much shorter time period - from ten seconds to two minutes - depending on the type of research and the complexity of the images.

Each image is processed by an algorithm and returned with the marking of potential problem areas. As indicated on the platform's website, its algorithms "are trained in combined methods, including using highly qualified medical markup."

By June 19, 2020, Third Opinion services are being tested in more than 20 Russian medical institutions. Medsi itself uses a startup-developed monitoring system for infected with COVID-19.

Medsi invested in Russia's first developer of medical AI services

Elena Brusilova, president of the Medsi group, told Kommersant that the network was the first medical company in the country to start introducing artificial intelligence services into practice. 

While artificial intelligence technologies in Russia are not widely developed, says the head of DSM Group Sergey Shulyak. At the same time, in the future, they should help partially abandon the work of a doctor, which will significantly reduce the company's costs, he said. Without cooperation with large medical centers, the development and testing of AI products is impossible, said the founder of the startup Anna Meshcheryakova.

The Medsi Group of Companies (through Medical Park LLC) and Platform Third Opinion LLC have also been co-owners of the legal entity Medsi - Third Opinion LLC since July 2019. GK owns a share of 51%, Platform Third Opinion - 49%.

Earlier, "Third Opinion" attracted 5 million rubles of investments from Sberbank. Natalya Magidey, Managing Director at SberX and Head of Sberbank's Acceleration Programs, also stressed that the bank will "strongly support" the Third Opinion starpap.

File:Aquote1.png
The effectiveness of the service is measured by the time that a medical professional saves. A doctor can spend from 1 minute (X-ray) to 1 hours to analyze medical images without using artificial intelligence (a difficult case on an MRI scan of the brain). "Third Opinion" gives the result almost instantly - the waiting time is from 10 seconds to 2 minutes, depending on the type of study, - said in "Medsi."[1]
File:Aquote2.png

As of June 2020, the Third Opinion company entered the markets of the Middle East, the CIS and the USA, was included in the list of finalists of the Russian-American accelerator of Sberbank and 500 Startups 2019.

Among the partners of the "Third Opinion" - Sechenov University, Moscow State Medical and Dental University named after A.I. Evdokimov, the company "Radiology of Moscow" and other organizations.

File:Aquote1.png
It is quite logical that the medical company is investing in developments in the field of artificial intelligence and computer vision - this is one of the most promising areas of diagnostics at the moment, - said Dmitry Gvozdev, General Director of Information Technologies of the Future. - Developments in this area are carried out by both narrow-profile companies - Medtronic, Medical System Screening, and those whose main profile is not related to medicine, for example, IBM, Microsoft, NVidia. Already existing technologies of this kind demonstrate significant effectiveness and are actively used in such areas as the treatment of cancer and many other diseases: it is safe to say that the future lies in the tight integration of AI technologies in the field of diagnostics. Already today, machine learning and artificial intelligence technologies make it possible to make completely unexpected discoveries - such as, for example, the applicability and effectiveness of drugs used to treat some ailments, to fight other diseases. In some cases, the speed and accuracy of recognizing a problem in medical images directly affects the effectiveness of treatment. Investments in this area are saved lives.
File:Aquote2.png

Obtaining a patent for a method for classifying blood cells using neural networks

At the end of April 2020, the Third Opinion company announced the receipt of a patent for a method for classifying blood cell types "using neural network technology"; the corresponding algorithm was trained on 120 thousand images and as a result recognizes more than 30 cell types with an accuracy of 82 to 99%.

The development was supervised, in particular, by Professor of the Russian Academy of Sciences Sergey Rumyantsev, Head of the Department of Oncology, Hematology and Radiation Therapy of the Russian National Research Medical University named after N.I. Pirogov Ministry of Health of Russia, as well as one of the leading experts in the field of computer vision Vadim Konushin.

2018: Plans to Launch a Neural Network Patient Image Analysis Service

Doctor Smart, a blockchain service for online health consultations, and the Third Opinion medical neural network in early February 2018 announced plans to jointly launch a medical image analysis service for doctors and patients. It is assumed that this feature will appear in one of the first versions of Doctor Smart, which will be launched in the first quarter of 2018. Read more here.