Several medical institutions in Russia have started using methods of detecting various types of cancer and other diseases. As Sergei Sorokin, Botkin.AI CEO and co-owner, told Invest Foresight, such institutions include Murmansk Regional Clinic, Tumen Regional Oncology Treatment Center, Ramsay Diagnostics chain. The 2015-born startup is thus becoming a practically operating business.
Founders and investors
Botkin.AI’s illnesses diagnostic program is an original development by Intellogic which was launched in 2015 and registered at Skolkovo innovations center in December 2017. According to Kontur.Focus service, Intellogic’s main shareholders at the moment are its CEO Sergei Sorokin (45.2%), a well known banker and Expobank owner Igor Kim (20%), and Primer Capital venture fund (20%). Oleg Bukhvalov and Ivan Drokin hold further 7.4% each.
Botkin.AI uses patients’ computer tomography (CT) or sets of X-ray images to detect cancer, and employs Artificial Intelligence (AI) which on the basis of neural networks can diagnosticate or even forecast diseases and submit clinical recommendations. At its current stage, the project only targets one area of medicine, oncology. It will nevertheless expand into other diagnostics areas. Botkin.AI can theoretically help detecting cancer at its early stage and anticipate the risks of the disease further development.
In July 2017, Intellogic got RUR 11 mio ($185K) investment from Primer Capital.
As Elizaveta Rozhdestvenskaya, Primer Capital’s COO commented on the deal, “We invested in Botkin.AI last summer, aiming toward completion of development of a cancer images recognition system. The development is by now at an implementation stage and will be employed in hospitals. But that is just the beginning of the process. Botkin.AI is in fact a new generation product. It offers Artificial Intelligence as a healthcare market service. So Primer Capital is pleased to support this dynamic project in Russia’s biotechnologies”.
The scope of investment to Botkin.AI by banker Igor Kim is not disclosed.
AI in hospitals
Intellogic has been cooperating with Blokhin Russian Cancer Research Centre and Moscow Institute of Physics and Technology. Yet to teach AI neural networks, Botkin.AI needs Big Data, while medics need diagnostics assistance. Last April, the company started cooperating with Murmansk Regional Clinic where Botkin.AI is incorporated into computerized retrospective review of chest CT images in order to detect lungs cancer at its early stages. It is also intended to assess risks of disease on the basis of electronic medical documents. Botkin.AI was also recently implemented at Tumen Regional Oncology Treatment Center, and last month it was launched at Ramsay Diagnostics. Sergei Sorokin admits though that cooperation with clinics is just in the process of establishment. Therefore statistics so far accumulated is yet insufficient to award reliable judgments on the medical method efficiency. Yet the data available suggests that Botkin.AI-based cancer detection is 95% which is a very high percentage.
Besides, Botkin.AI is employed by some ‘international pharmaceutical company’ for analysis of information on medications’ use efficiency. Sorokin declined to disclose the company name though.
Sorokin did not specify the fees Intellogic collects from the clinics for their use of Botkin.AI. He nevertheless hopes to see the project break even within a year.
Why Botkin.AI took risks of entering markets
Botkin.AI’s most prominent competitor is IBM Watson. In various clinics in the US and Thailand Watson for Oncology program is already employed by medics for diagnostication purposes.
“It is much cheaper to implement our platform than IBM Watson. Besides, we use a different technology”, Sergei Sorokin says.
Botkin.AI is a patented AI-based mathematical model for patient representation. The model is based on three types of data, namely, graphic images, structured data and unstructured medical data. It analyses ECG, CT, MRI, ultrasound, medical test results, genomic data, patient’s health records. In other words, Big Data analysis is performed to review all data collected by doctors. When making its diagnosis, IBM Watson, analyses not just patient’s medical data, but TBytes of the recent research on the subject. Its diagnosis is thus based on the totality of all data available. No oncology specialist is able to compete with AI in swiftness of information collection and analysis. IBM Watson, for instance, spent merely 10 minutes to analyse a patient’s genome and come up with a treatment plan, whereas doctors needed 160 hours for that. The price of the service for a patient ranges between $200 and $1K.
Botkin.AI does not provide the information on the swiftness of its operation.
Among all Russian companies which strive to diagnosticate diseases by employing AI (Third Opinion project should in the first turn be named), Botkin.AI is the first to come to the market with a product ready for implementation.
By Natalia Kuznetsova