A living lab: How medical data science is developing in the Amsterdam Area

The word unprecedented might have been overused during the coronavirus crisis, but it does succinctly sum up the global impact of this deadly disease. What has happened over the last few months has been exceptional; the response of people all over the world extraordinary. And nowhere is that impact felt more keenly than in the life sciences and health sector.

While frontline workers are caring for those affected by the disease, medical professionals are also working tirelessly behind the scenes to solve challenges that can help contain the outbreak and limit its still uncharted consequences for the world economy. These challenges are numerous: health risks need to be identified more quickly and efficiently; algorithms need to be developed and refined to improve diagnostics and recommend treatments; quantities of hospital beds need to be constantly managed; and the time needed to develop, test and distribute effective vaccines has to be shortened.

Life Sciences and Health in the Amsterdam Area

For all these challenges, the Amsterdam Area’s life sciences and health sector is using artificial intelligence (AI) innovations to provide the right answers – with the city region as its smart lab. Amsterdam’s data scientists and medical professionals are using AI to deliver solutions and create new partnerships that increase our resilience to the virus. Happily, this work is also improving the efficiency and performance of the sector as a whole.

The Amsterdam Area: a medical data pioneer

Asked to explain Amsterdam’s pole position in developing AI solutions to combat the crisis, Jeroen Maas, Challenge Lead for Life Sciences & Health at the Amsterdam Economic Board, says: “The infrastructure for both medical and biomedical science and information technology is top level. This attracts a continuous influx of talent and an international-oriented network of data-driven companies.

“Also, we were early adaptors and recognised the potential of AI at an early stage. Add the ecosystem of specialised cure and care providers and hundreds of companies that specialise in health-related products and services, and you understand why there are so many innovations in health and life sciences spurred by a smart use of AI and machine learning here. Especially in the field of patient-data.”

Jeroen Maas

Jeroen Maas

Mark Hoogendoorn, associate professor AI & Health of the VU University Amsterdam, adds: “Because of the growing sense of urgency since the outbreak, there is a boost in demand and an increasing willingness to share data. AI is no longer a promise, it is acknowledged as a robust tool that enables technology to make better-informed decisions and mitigate complicated risks now. Not only in healthcare but in other domains, too.”

Though it might seem too good to be true, there are plenty of case studies and businesses that offer evidence of the Amsterdam Metropolitan Area’s credentials as a leading AI hub. Take Amsterdam’s Pacmed, which is already using AI to help medical professionals make optimal use of ICU beds around the Netherlands. If a patient is kept for too long (or for not long enough) in intensive care, it can cause serious problems for their recovery. Pacmed’s machine learning model helps doctors to make a discharge decision, reducing the chances of a patient being readmitted or developing complications further into their recovery.

Pacmed's team have won awards for their work in AI

Due to their expertise in medical data, Pacmed’s award-winning team were also invited to collaborate with virologists and epidemiologists from Amsterdam UMC and Vrije Universiteit Amsterdam data scientists to design a solution to combat shortage of IC beds for COVID-19 patients. The result is a successful tool that can help predict demand in ICUs and prevent hospitals from becoming overburdened. The data the model uses includes the reproduction number (a way of rating a disease's ability to spread), regional demographics, clinical records and available ICU capacity in the Netherlands and neighbouring countries. Next, the tool will be part of the pandemic dashboard that will help the Dutch government to make decisions to manage lockdown measures.

How algorithms can help crack the corona problem

The Netherlands Cancer Institute (NKI) also took the lead when the coronavirus outbreak happened. It almost immediately urged more than 30 hospitals around the EU to submit radiographs of lungs to help create an algorithm that recognises the various symptoms of the disease. This way, a new dimension was added to help diagnose people with coronavirus. Now, its team is working to includes CT scans which can help provide more detailed information.

Netherlands Cancer Institute (NKI)

The Amsterdam UMC Imaging Center is a pioneer in medical imaging. It is home to the most advanced equipment available and the latest innovations in diagnosis-support. This allows radiologists and physicians who come to work at the centre from all over the world to better examine the anatomical structures and physiological, metabolic and molecular processes of the human body. Here AI is being used to find unseen patterns that can help us better understand the complexities of the virus (or other diseases). It also helps to speed up the time needed for clinical tests. As the world anxiously awaits a breakthrough in the search for a proven effective vaccine or virus inhibitor, this work is of the utmost importance.

Why privacy is key to unlocking the benefits of big data

One of the biggest stumbling blocks to unlocking the benefits of big data is, with good reason, privacy. In the face of COVID-19, rules have been streamlined so data can be shared more easily. Privacy is paramount, and all patient data is anonymised and kept on a highly secure server (those who are not comfortable sharing their data can also opt-out). But, industry leaders say that as people take more interest in how their data is being used they’re also more willing to share it for the right reasons. In addition, private sector firms are sharing their data more openly for the greater good, instead of treating it as intellectual property which external parties pay for to use.

Amsterdam UMC Imaging Center

Amsterdam UMC Imaging Center

Amsterdam’s work to unite data scientists and doctors

Home to world-leading experts, research institutes and public and private academic labs, work has also been quietly going on in the Amsterdam Area to team medical professionals up with data scientists. This mission has helped spawn several collaborative projects worked out over coffees and pizza (yes, really), and in the halls of Amsterdam’s universities and hospitals. And it all started with stakeholder and business leaders realising that these different disciplines needed to share their experiences and insights to help translate medical data to clinical settings.

Jeroen Maas says: “There is no such thing as a formula for success, but our triple helix approach certainly helps." That method refers to a three-way collaboration in which all participants - governments, academia, companies, health providers, insurers and patient-organisations - work closely together to explore how AI can to create a smarter and healthier society. “In short, the local and national governments set and maintain the rules and create the best possible conditions to enable academics to research, discover and evaluate interventions independently and the private sector to transform discoveries into valuable products.”

AI in the Amsterdam Area

Hoogendoorn thinks that the current AI solutions for the life sciences and health sector help us in three ways. He says: “In diagnosing patients and assessing health risks faster and better, and in flattening the curve. But, although AI is a powerful tool that can help us to ease the effects of a pandemic, it doesn’t solve everything. Politicians and business leaders need to understand this.”

The coronavirus pandemic has made an unparalleled impact on the world as we know it, and it is not easy to look past the tragedies that are unfolding in homes and hospitals each day. But, one positive is the inspiring way medical professionals, data scientists, entrepreneurs and even patients are all helping to play a part in developing new methods to help others. Unprecedented, yes, but the work being conducted around the Amsterdam Area is also important, remarkable and capable of paving the way for creating new methods to help keep people healthy.

Read more about the Amsterdam Area's life sciences and health sector.