In the future, doctors and nurses will be less likely to miss signs of lung or heart failure. A new algorithm can predict a number of critical illnesses, which will give doctors peace of mind and relieve pressure.
This is the opinion of Ulf Hørlyck, senior consultant at the Emergency Department at the Regional Hospital in Horsens.
- The algorithm is the most innovative thing I've seen in a long time - from an employee safety perspective," says Ulf Hørlyk.
The AI algorithm +Priocritical has just been tested at Horsens Regional Hospital. The researchers behind the algorithm state in an article that the algorithm clearly outperforms traditional methods for diagnosing, for example, sepsis (blood poisoning). In addition, the thesis is that the algorithm can detect critical illness earlier than traditional methods, which will mean a high quality improvement for patients.
The algorithm is incorporated into a smartphone app. The app constantly collects data, and the employee receives a notification if the data shows an increased risk of a number of critical illnesses, explains healthcare engineer Simon Meyer Lauritsen.
- As soon as a new blood sample is analyzed or a vital parameter is measured, the new data is collected so that the risk assessment is constantly updated. The app also provides the objective data so the doctor or nurse can see the reason why the notification is sent, says Simon Meyer Lauritsen.
+Priocritical is a further development of Simon Meyer Lauritsen's PhD thesis. He has based his algorithm development on historical data from the data project CROSS TRACK in a business PhD program in close collaboration with the company Enversion A/S, Horsens Regional Hospital and Aarhus University.
Gitte Kjeldsen, Danish Life Science Cluster, is the project manager for the AI Signature project where the algorithm was developed.
- We are excited to leverage the clinical data and further develop Simon's research-based work into a concrete solution. This algorithm brings a new perspective to AI-driven decision support tools and opens up new, broader solutions in early detection with AI," says Gitte Kjeldsen. She adds:
- One of the major challenges in healthcare is to transform academic knowledge into concrete, usable tools for healthcare professionals, and we have succeeded in doing so to the benefit of all.
The partners have just completed a three-week design sprint with +Priocritical with interactions at the Regional Hospital in Horsens. The Emergency Department, Medical Department and Anesthesiology Department contributed with medical, nursing and management skills.
The development work has also involved anthropologists, UX designers, researchers and data engineers.
The feedback from healthcare professionals has been hugely positive.
This interdisciplinarity should be the basis for all healthcare innovation, and in the development of AI solutions, interdisciplinarity is simply indispensable, says Simon Meyer Lauritsen.
- Feedback from the clinic is indispensable for us technicians. We need to know what features they need and how a device adds the most value. Researchers and healthcare professionals work in two very different scenarios, so we need to come together to understand each other's challenges. I'm proud that my geekiness has resulted in a workable solution, but we've only gotten this far because of the support of healthcare professionals," says Simon Meyer Lauritsen.
The next steps in the project will be to customize the algorithm and test it in the clinic with real-time data, so work is underway to create the optimal framework for this.
