An AI Research Tool That Solves Medical Data Siloing
By: Jamie Bussin
A data silo is a group of raw data that is accessible by some but isolated from all. When it exists in a company it results in a severe lack of transparency, efficiency and trust within. When it exists in the context of the medical research community it also results in inefficiencies, costs and missed opportunities to develop drugs and treatments. We live in a digital society. The data that is being created has the potential to help us solve complex health issues, if it can be properly harnessed. But if that information exists in different formats and configurations, is voluminous and uncoordinated, and perhaps isolated from other relevant data, its value can not be capitalized upon.
In episode #244 of The Tonic Talk Show/Podcast I spoke with Aaron Leibtag, the CEO of Pentavere Research and Jefferson Tea, the Vice President of Medical and Scientific Affairs at Takeda Canada regarding Pentavere’s AI research tool which tackles the problem of information siloing in medical research and how, in winning an open competition for funding by Takeda, their goals are forwarded. This is a digest of that conversation.
Takeda Canada is part of Takeda, a global pharmaceutical company that develops medicines that primarily treat cancer, diseases of the gut (such as Ulcerative Colitis and Crohn’s Disease), diseases of the brain and rare diseases (those which affect less than 1 in 2000 people). According to Mr. Tea, most diseases do not have a cure, and so it is important to Takeda to invest in research to bring new medications to patients and contribute to society.
Mr. Tea believes that; “Healthcare innovation is critical to our business. When we think about the different diseases that patients are dealing with there is a huge unmet need. So from a business standpoint there is a need to speed up the research process -to bring these new solutions, these medicines, to patients who are suffering and to improve outcomes. We can only do that with health care innovation; using new technology, artificial intelligence and data analytics to speed up research. It goes without saying that we can’t do that alone. We need to develop relationships with academia, biotech companies, and start-up organizations to bring medicines to the patients.”
Earlier this year Takeda Canada held an innovation challenge inviting health tech startups to submit applications to support enhanced patient care in rare disease conditions. Takeda wanted to find new thinking and companies that traditionally wouldn’t be in a position to access big pharma research funding, and also to identify ways to bolster innovative solutions to support rare disease patient care. Takeda asked these companies to explain how their technologies can solve healthcare problems, specifically patients suffering from rare diseases.
The winner of the competition, securing $200,000 in health tech funding and lab support from Takeda, was Pentavere Research, an AI clinical discovery company. They developed and proved an AI engine “Darwen” that accelerates and unlocks knowledge from huge amounts of clinical information that exists, but is otherwise not used, to solve health problems. Darwen discovers high value clinical information contained in electronic health records, physician notes, pathology reports, physician transcriptions and other digital sources.
As Mr. Leibtag explains; “When we think about what is in our electronic health records – that documentation, that information that we go through with our doctor, our diagnostics, all of our comorbidities, all other aspects of our information that is out there, it is not used today at scale to be able to understand what is happening in the real world, or to understand how knowledge can be used to accelerate discovery and diagnosis. We’ve developed a tool that solves that problem.”
The siloing of medical data is a key problem. It also speaks to why Mr. Leibtag started Pentavere and what their process is in terms of bringing the company to life. Pentavere was started out of a tragedy. Mr. Leibtag’s co-founder, Steven Aviv, was a technologist who spent 20 years architecting global systems to uptake data for financial companies to make money in real time. Mr. Aviv’s mother went into the hospital for what should have been a routine procedure. But an important note about a medication that was required post surgery went missing by her care team – and she never made it out of the hospital. Messrs. Leibtag and Aviv embedded themselves for two years at a major hospital system in Canada to learn firsthand about all the complexity around data silos. They tried to solve the questions of how to get access to data, and how to contemplate privacy by design – those things that are just as important as the AI algorithms that extract the data; to be able to unlock and unleash the information so that companies such as Takeda could accelerate early diagnosis of rare diseases.
Regarding privacy issues pertaining to medical data Mr. Leibtag says; “We not only have to understand it, we have to build the solution into our process, and not compromise on them. If an AI algorithm doesn’t take into account privacy, governance and ethics in the process, then you’re not going to solve the fundamental problem that we need to solve; that is how to unlock the information to help improve the life of Canadians with the best science available.”
To that end one of the first major decisions Pentavere made was to hire the former chief privacy officer at University Health Network. Thus Pentavere was able to understand privacy, and demonstrate to the stakeholders that it understood the obligations and built trust. As a result, in many health systems across Canada, Pentavere is the first commercial AI engine that has been deployed and gained access in the most responsible way to unlock and unleash information and publish that information in the public domain around high value clinical outcomes that can be used to improve care and save lives.
Jamie Bussin is the Publisher of TheTonic Magazine and Host of TheTonic Talk Show/Podcast. For more information about Takeda Canada visit https://www.takeda.com/en-ca/
For more information about Pentavere Research visit https://pentavere.ai/