Unlearn.AI nabs $12M to build “digital twins” to speed up and improve clinical trials

Twins have long played a role in the world of medical research, specifically in the area of clinical trials, where they can help measure the effectiveness of a therapy by applying a control to one of a genetically-similar pair. Today, a startup founded by a former principal scientist at Pfizer, which has developed a way of digitising this concept through the use of AI, is announcing some funding to further its efforts. Unlearn.AI, which has built a machine learning platform that builds “digital twin” profiles of patients that become the controls in clinical trials — is announcing that it has raised $12 million in a Series A round.

The round is being led by 8VC with previous investors DCVC, DCVC Bio and Mubadala Capital Ventures also participating.

The startup’s DiGenesis platform is first being applied to neurological diseases, specifically Alzheimer’s Disease and Multiple Sclerosis, where effective treatment options remain an elusive goal and it has been hard to build clinical trials around patients with already-impacted health.

Although Unlearn.AI is not working on anything close to medicines related to the COVID-19 pandemic, it’s a timely reminder of why improving clinical trials is important. We’re now in an urgent race to find vaccines and treatments for this new virus, and that highlights the need for more efficient approaches to trials, and that is an area where AI could prove to be a boost.

Unlearn does not disclose who its commercial partners are today, nor how far they’ve come with active, live trials. It was in discussions with regulators before the coronavirus outbreak halted everything. The funding will be used to inch closer to commercial rollouts, it seems.

“This new financing marks an important milestone in our growth and will contribute to the significant progress we are making with regulators and with our commercial partners, who are already running studies with Digital Twins and demonstrating their value in generating robust evidence and increasing the potential for trial success,” said Charles K. Fisher, Ph.D., founder and CEO of Unlearn.AI, in a statement.

“Clinical trials are facing a number of persistent challenges that have only been exacerbated in recent weeks. With support from our forward-thinking investors and industry partners, we are excited to continue growing our exceptional team and advancing the science behind our first-of-its-kind Digital Twin approach.”

Fisher’s background is one that falls squarely at the nexus of technology and medical research. In addition to time spent as a principal scientist at pharma giant Pfizer, he has also worked at Leap Motion, and those roles followed years of studying and researching biophysics in academia.

Unlearn approaches the idea of building these so-called digital twins as a classic machine learning problem, using “clinical trial datasets from thousands of patients to build the disease-specific machine learning models used to create Digital Twins and their corresponding virtual medical records.”

These are more than simple medical profiles: they match people according to demographics, lab tests and biomarkers. The idea is that by building AI-based twins, there is less of a need to find similar actual pairs of people — actual twins, even — to run tests and controls.

Unlearn has been working on its platform since 2017, but the use of twins (and the pair’s very close genetic makeups in medical research) to track pathology and treatments goes back decades, and interestingly one of the novel coronavirus tracking apps that has seen some strong traction was borne out of a long-term twins study run out of Kings College Hospital in London working with Stanford and Massachusetts General Hospital in the US.

The growth of using AI to build “people” to run the effects of drugs also follows a much bigger theme of using computers and algorithms to test and create chemical combinations and therapies that would have in the past taken much longer, and cost much more, to run out manually. (Another example of where this is being applied is in the world of product development, where consumer goods companies are using AI platforms to formulate new soaps and other goods.)

“Unlearn’s pioneering use of Digital Twins will limit the number of patients that need to go on placebo while also reducing overall trial enrollment time,” said 8VC Principal, Dr Francisco Gimenez, in a statement. “As investors at the intersection of healthcare and technology, we’re passionate about companies that pair cutting-edge computational techniques and innovative business models to meaningfully improve patient care. 8VC is excited to partner with Unlearn to bring about the biggest change in the drug approval process since the RCT.” Gimenez is joining the board of the startup with this round.