Copyright infringement

Unlearn.AI, a startup developing a ‘digital twin’ service for clinical trials, raises $50M –

The concept of digital twins – the digital representation of human beings built on computer models – is gaining ground in the private and academic fields of health research. Predictable imaging technology, some experts say that digital twins are able to improve health care by assessing health risks before the disease becomes a symptom, helping doctors decide, for example, when (and ) to interfere.

The future in which doctors can limit the impact of all possible digital twin treatments to patients to determine the most effective course is really promising. That may be the reason Learn nothing.AI, a startup company announced today that it has raised $ 50 million for the Series B funding round, launching clinical trials. Unlearn’s digital twin product restores patient characteristics to experiments to enable what the company claims to be small, fast-paced studies, built on a combination of AI and historical data.

“We use the data combined with a large number of previous clinical trials. Our product is not an AI model – it is a clinical trial,” CEO Charles Fisher told TechCrunch via email. “Ensuring the development of the vaccine [during the pandemic] It means that every journalist and most informed client is painfully aware of the need to expedite patient tests while doing it safely … [While there] are other companies interested in using real-world data to make experiments faster or better, there are no direct competitors with a direct legal route to use their technology [late-stage] medical tests. “

Unlearn was founded in 2017 by Fisher, Aaron Smith and Jon Walsh – all physicists by training. The trio met while working together on Leap Motion, a now-defunct startup that develops motion sensors on hand-held tables and extended headboards.

Fisher, Smith and Walsh wanted to create a service that could analyze historical clinical trial data for patients to build “disease-specific” machine learning models, which could be used to create digital twins with medical records. straight. The digital twin records will be longer – that is, add data over time and to all systems – and will cover statistical data, routine test results and biomarkers that look similar to real patient records on trial. health.

“[Our] The goal was not to expedite patient experiments – it was pure research on machine learning. But [I] He had a pharmaceutical background and it soon became clear that there was no investment in machine learning such as pharmaceutical development technology, “said Fisher.[Unlearn] developed in integration[s] with the pharmaceutical industry. “

Today, Unlearn works with pharmacists, biotechnology companies and academic researchers to create digital twins for each patient in the clinical trial. Fisher said The therapeutic effect can be quantified accurately after correcting the results obtained from digital twins.

Unlearn’s capabilities are reportedly enough to convince three companies to get involved in its product studies, although Fisher only wanted to name one: Merck KGaA, Darmstadt, Germany (separate pharmacy from Merck). Merck KGaA uses Unlearn to include predictive data on digital twins in its randomized trials, which hopefully will enable smaller control groups and produce evidence “appropriate support for regulatory decisions in the vaccine pipeline.” , “according to Fisher.

If Unlearn digital twin technology works and is also advertised, it could be a god for the healthcare industry that has long been forced into high costs and logistical challenges related to medical experiments. According to a 2018 study from Johns Hopkins, clinical trials support The U.S. Food and Drug Administration’s approval of new drugs has a median value of $ 19 million. These medical tests – which occur in several, several months – can last for years (six to seven) on average) and face unexpected challenges such as a lack of knowledgeable participants and a change in protocol.

But a few education raise questions about limiting digital twin technology, as well as its potential vulnerability to biased data. One close paper she noted that bias – resulting from, for example, the reduction of Black patients in clinical trial data – could affect the accuracy of predictions using digital twins.

Fisher disputed the notion that Unlearn technology could lead to erroneous decision-making, pointing to draft opinion a European Medicines Agency (EMA) showing that digital twins can be used for basic analysis in phase 1 and phase 2 drug studies. (EMA is approximately the same as the pharmaceutical component of the U.S. Food and Drug Administration.)

“Tthe question is whether there can be bias in patient testing using this technology. It would be mathematically impossible. ” Fisher said. “[Moreover,] Unlearn only uses compact data and cannot access private data. “

With the new capital, bringing the total Unlearn total to $ 69.85 million so far, the company plans to double the number of 40 heads and expand operations to new disease sites.

“The medical testing technology industry has one big problem: Pharmaceutical companies are skeptical of new technologies,” Fisher said. “The biggest challenge is to build evidence to convince them that the new methods will bring value while still providing evidence that can be used in the regulatory process.”

Insight Partners participated in Unlearn’s Series B, which also saw the participation of Radical Ventures and existing investors 8VC, DCVC, DCVC Bio and Mubadala Capital Ventures.