My Mission to Find Stroke Treatments Using AI

By Andrea Derix, R&D Global Program Management Thrombosis Portfolio at Bayer

Cardiovascular disease has touched those close to me, like it does millions of people globally. It affects all faiths and cultures. And unfortunately, it often leads to severe cardiovascular (CV) events like heart attacks and strokes. That is why I have spent so much of my life as a researcher looking for innovative ways to improve prevention and management of CV diseases, especially in the area of stroke.

Many years ago, a close friend of mine experienced a stroke in his early 60s, and I saw how debilitating it was to his career and personal life. He was no longer able to carry on with his professional and private life in the way that he had been used to because the stroke had substantially reduced his mobility. It was very emotional to see how one event could completely change a person’s life. Depending on severity of stroke, not all functions will improve, and preventing serious health events can really make a difference to people’s lives.

Cardiovascular disease is the leading cause of death worldwide. The World Health Organization reported an estimated 17.9 million people died of cardiovascular disease in 2016–85 percent of these deaths were due to stroke and heart attack.

When I joined Bayer 15 years ago, my mission was to continue to find the best medicines to treat and prevent cardiovascular disease. Today, I’m focused in the area of stroke and heart diseases. We have been successful with our development in the past with drugs that help to prevent strokes in patients with have atrial fibrillation. Still, there’s a remaining group of patients who do not have atrial fibrillation, but have different underlying risks. Those especially in the elderly population. There’s also a group of people who need preventative measures to stop a second stroke from happening.

It is becoming more and more difficult with classical research methods to achieve the results we want.

This has led me to the area of artificial intelligence — or AI — which I believe will point us to the next generation of medicines. AI may open new doors for the next level of development. It could also add value through decreased research and development risk and reduced costs through more targeted and smaller clinical trials.

Together, Bayer and the Broad Institute of MIT and Harvard, are pursuing the goal of developing AI algorithms to identify patients with a high risk of cardiovascular diseases, including heart failure and recurrent stroke. The Broad Institute’s connection to Partners HealthCare and its network of provider institutions allow for the possibility to scientifically analyze large sets of anonymized patient electronic records. We are seeking to learn more about the various causes of disease, what a patient reports over time — in terms of incidents, disease factors, treatments before experiencing a stroke. With that knowledge, we hope we can identify patients who can benefit from a customized approach to therapy. We need to look earlier at retrospective data to understand what we can do as a preventative measure for patients.

Over the next few months, a team of data scientists, together with biologists and medical doctors will start working together to analyze and explore the data with different artificial intelligence methodologies. The use of such technologies like machine learning will enable us to work with rich, diverse and large datasets.

My vision is to gain new insights into the diseases, that can then translate into more effective and precise drug development for cardiovascular disease patients. The hope is that, thanks to AI, eventually markedly fewer people will suffer the way my good friend did.