Skip to main content
European Commission logo
English English
CORDIS - EU research results
CORDIS
CORDIS Web 30th anniversary CORDIS Web 30th anniversary

Perioperative infection prediction

Article Category

Article available in the following languages:

Artificial intelligence helps predict – and mitigate – post-surgical infections

A Dutch medical technology company is using advanced machine learning to quickly and accurately predict post-surgical infections.

Every year, over 50 million people undergo inpatient surgery in Europe. Out of these, an estimated 25 % will get an infection within 30 days after the surgery – that’s nearly one out of every four post-surgical patients. Beyond the personal pain and suffering this causes, there’s also a cost of about EUR 10 000 per patient. Despite advancements in biomarkers, risk scores, early warning devices and preventative methods, the risk of catching an infection after surgery remains high. This is why Healthplus.ai, a Dutch medical technology company, is developing PERISCOPE. “PERISCOPE is an advanced machine-learning algorithm that reuses available electronic health record (EHR) data to predict post-surgical infections 5 days prior to the average medical team diagnosis,” says Bart Geerts, CEO at Healthplus.ai. Although the tool already has an 88 % accuracy level, the ultimate goal is to achieve 95 % accuracy. With the support of EU funding, not only is the company well on its way to reaching this goal, it’s also a big step closer to bringing PERISCOPE to market.

Staying one step ahead of the competition

With PERISCOPE’s technical feasibility already established, Healthplus.ai sought to assess potential channels for delivering the tool in a safe, affordable and scalable manner. This meant identifying third-party vendors and partners who exhibited sustainable business models. To ensure a smooth market entry, the team also investigated legal and regulatory issues at both the EU and Member State levels. “Through these interviews, we have built – and maintained – meaningful relationships with hospitals across Europe, as well as large international EHR vendors, most notably Cerner and ChipSoft,” explains Geerts. The team also conducted detailed interviews with stakeholders, including end users, hospital administrators, insurance companies and EHR vendors – among others. Not only did these conversations reinforce PERISCOPE’s market potential, perhaps more importantly, they unveiled potential issues and areas for improvement. “This market research let us learn about our product and identify barriers to entry from a commercial, technical and clinical perspective,” says Geerts. “We were able to redefine our product, roadmap and partnerships with hospitals, EHR vendors and other key entities.” Through the project, Healthplus.ai also expanded its R&D partnerships and grew its datasets and clinical input. “These are essential steps to staying one step ahead of the competition and clinical standards, thus giving us additional competitive advantages,” adds Geerts.

A new era in healthcare

Thanks to support through EU funding, PERISCOPE is set to redefine post-surgical care. “The adoption of artificial intelligence (AI) in the healthcare sector is still limited, especially where EHR data is reused,” says Geerts. “This project represents the start of a new era in healthcare, one which uses AI and reuses EHR data to make surgical care more proactive.” The company is currently finalising the product and preparing for its commercialisation. To do this, it is looking to secure the second phase of the SME instrument and Eurostars grants, in addition to venture capital investments. The team is also advancing its clinical evidence and business case, along with securing CE marking and FDA approval – both of which are critical to bringing PERISCOPE to the mass market.

Keywords

PERISCOPE, artificial intelligence, AI, post-surgical infections, medical technology, machine learning, inpatient surgery, electronic health record, EHR, Healthplus.ai, Cerner, ChipSoft, healthcare

Discover other articles in the same domain of application