Every year, 150,000 major operations are performed in Danish hospitals, and around 30% of these lead to serious complications. As a consequence of complications, patients need to be hospitalised for longer and, in the most severe cases, they suffer permanent discomfort or can even die. Similar risks apply for the approx. 50,000 patients in Denmark suffering from severe pulmonary disease (COPD); a disease which leads to thousands of admissions every year.
Using novel technology, a new project will revolutionise monitoring of high-risk patients admitted to Danish hospitals. By means of a number of small wireless sensors placed on the patient’s body, the technology continuously monitors about ten different parameters, for example blood pressure, oxygen saturation, heartbeat and respiration.
The measurements are sent wirelessly to a central computer constituting the ‘brain’ of the new monitoring system. Using artificial intelligence (AI), the central computer reads, analyses and interprets the vast amount of data, and immediately notifies hospital staff if complications are developing and the patient needs to be attended to.
Strong focus on monitoring
“In Denmark and other countries, recent years have seen the introduction of systematic patient monitoring by measuring blood pressure, oxygen saturation, etc. 2-3 times a day. Substantial personnel resources go into monitoring, without this leading to fewer complications or deaths. One explanation could be that complications may occur during the 8-12-hour intervals between measurements, and since the patients don't always feel the complications themselves, the complications are often detected too late for the health staff to intervene in time. Our new system will address this problem,” said Consultant Eske K. Aasvang, head of the research unit at the Department of Anaesthesiology at the Rigshospitalet Abdominal Centre.
Eske K. Aasvang is one of three experts behind the project. The other two are Christian S. Meyhoff, Consultant and Head of Research at the Department of Anaesthesia and Intensive Care, Bispebjerg and Frederiksberg Hospital, and Associate Professor Helge B.D. Sørensen from DTU Electrical Engineering, both of whom are responsible for the technical-scientific part of the project. The project includes measurements from a total of 5,700 patients selected in collaboration with Odense University Hospital and Aarhus University Hospital.
Monitoring based on artificial intelligence
The large amounts of data resulting from the 24-hour monitoring will be read and interpreted by means of artificial intelligence, i.e. a number of bio-medical algorithms and data models enabling the future monitoring system to assess whether the patient data are normal, or whether there is a deviation – a so-called micro-event - which could develop into a complication.
“For the system to work, as physicians we need to ‘teach’ the algorithms when to react because a parameter is measured as deviating from the normal values. We will also teach the algorithms about which complications a given deviation may indicate, for example pneumonia. Once all these data have been fed into the monitoring system, the system will be able to detect or even predict a complication and send a warning to the nurse, or straight to the physician, in the case of a major deviation, e.g. a blood clot in the heart,” explained Consultant Christian S. Meyhoff, Bispebjerg and Frederiksberg Hospital.
Enhancing the knowledge of physicians
The three experts behind the new monitoring system, WARD (Wireless Assessment of Respiratory and circulatory Distress) will make the system as skilled as an experienced medical specialist.
“Knowledge will also go the other way so that physicians know even more than they know today, because the artificial intelligence of the monitoring system will be able to identify correlations and patterns that physicians are not currently aware of. For example, the combination of changes in oxygen saturation, blood pressure and a range of other parameters during a period of 24 hours could be associated with complications of which we have not previously been aware, but which become evident through the combination and automatic interpretation of large data sets,” said Helge B.D. Sørensen from DTU Electrical Engineering.
Over the coming years, the project group expects to present a complete monitoring system that can be implemented at hospitals in Denmark and the rest of the world.
Innovation Fund Denmark has invested DKK 18 million in the project, and the other project partners will contribute another DKK 12 million. Project funding has been provided by the Danish Cancer Society, Radiometer, Bispebjerg and Frederiksberg Hospital, Rigshospitalet, the Technical University of Denmark, Isansys and A.P. Møller Fonden.