With a solution for predicting infections after surgery, Healthplus.ai, PW Consulting and SAS have obtained a grant of 1.27 million euros from the REACT-EU recovery fund. The consortium will use the funding to scale and cost-effectively deploy PERISCOPE™, the artificial intelligence solution of Healthplus.ai, across multiple hospitals. PERISCOPE makes it possible to predict and reduce the number of post-surgical infections within hospitals.
Infections after surgery are a major problem for hospitals. On average, 1 in 5 patients is affected. The longer recovery time and additional treatments involve enormous costs, on average about 10,000 euros per year. With current resources, the diagnosis is now often made around day 5 after the operation, which means that valuable time is lost, and recovery takes longer.
AI model for predicting post-operative infections
Based on this information, the Dutch startup Healthplus.ai developed PERISCOPE, an artificial intelligence model that can predict post-operative infections much earlier. The model was tested at Leiden University Medical Center (LUMC). The initial results are promising. The model currently has an accuracy of 91% in predicting infections within the 30-day period after surgery. Also, the algorithm seems to be reliable for all adult patients and almost all surgical specialties.
Bart Geerts, founder of Healthplus.ai, says: “Now that we have proven with LUMC that the algorithm performs well, we want to make PERISCOPE available to many more hospitals. This way, we can contribute to minimizing the damage caused by an infection and reduce the time that patients have to spend in hospital after an operation. We are therefore particularly pleased with the allocation of the REACT-EU funding, so that we can really get started.”
The biggest challenge is the applicability
The biggest challenge is to apply the solution within all regulatory requirements, because PERISCOPE is a (software as a) medical device. However, each hospital has its own characteristics in terms of geographical location, patients, employees, processes and technology. Therefore, it is challenging to deploy PERISCOPE in a scalable, secure and cost-effective manner at different hospitals. The goal of the project is to conduct a clinical study in which PERISCOPE will be applied in multiple hospitals. Next, CE certification will be obtained.
Joost Huiskens, Industry Expert Healthcare at SAS, says: “The biggest challenge is not the development of artificial intelligence but its deployment. Our goal is to make PERISCOPE as accessible as possible to every hospital using SAS Viya. We strive to do this quickly, scalable and within all regulatory requirements. This will result in more people being helped and a reduction of the time patients spend in hospital. And all this will contribute to more efficient and cost-effective healthcare. We are extremely proud that we from SAS can contribute to these developments.”
To reduce the impact of post-operative infections on a large scale, it is important that all hospitals can implement PERISCOPE easily and with a minimum of risk. The architecture that is designed for this purpose ensures that the technical implementation of PERISCOPE is simple, while also ensuring that the specific requirements of a hospital are met. The necessary future development of PERISCOPE requires a substantial investment. To realize this, a request was made to the REACT-EU fund.
There are still few successful examples of AI applications in healthcare, especially when it comes to the Medical Record and not radiology or pathology. Offering a scalable, proven AI solution to realize impact for a large group of patients in the Netherlands and soon the EU is unique and a strong personal motivator for the team of Healthplus.ai, PW Consulting and SAS.