New technology identifies COVID-19-infected patients

Insight

New technology identifies COVID-19-infected patients

Researchers and clinicians from Copenhagen University and DTU’s Department of Health Technology have developed a ground-breaking monitoring system, WARD, which uses artificial intelligence and wireless technology to monitor patients even when they are not hospitalized.

When

03.04.2020

As the coronavirus (COVID-19) pandemic spreads across the world, The Novo Nordisk Foundation has fast-tracked grants for projects that can help fight the COVID-19 pandemic and support the infected, and the WARD project recently received such a grant. The WARD technology will now be implemented to help prevent the spread of coronavirus infection, warn of critical deterioration and create computer models to predict the development of Covid-19.

Under normal circumstances, a hospitalization would be the safe choice when a patient is at risk of developing serious, life-threatening illness, but with COVID-19, nothing is as it used to be. The lack of protective equipment, and the risk of disease transmission through physical contact between health professionals and patients represents a unique challenge for the health care system, and wards will be under pressure for months to come. New methods based on artificial intelligence are the obvious solution. This technology is about to be introduced in a number of Danish hospitals, and COVID-19 patients will be among the first to benefit.

From major surgery to COVID-19

WARD-COVID-19 is an adaptation of the ongoing project WARD (Wireless Assessment of Respiratory and circulatory Distress), which is designed to assist medical staff by improving monitoring of patients. The system has received funding from the Innovation Fund Denmark, and was originally aimed at high-risk patients (e.g. those with Chronic Obstructive Pulmonary Disease or heart disease) and patients who have been through major surgery. Last week, the Novo Nordisk Foundation granted DKK 3.9M to the WARD team at Rigshospitalet, Bispebjerg Hospital and the DTU to implement an early version of this technology to reduce the need for close contact between health care staff and hospitalized COVID-19 patients. Many COVID-19 patients are very unstable and show rapid and unpredictable deterioration without the usual clinical warning signs, so health care staff have to check them very frequently. To avoid contamination, the healthcare staff wears personal protective equipment, which is a limited resource and time consuming for staff to use.

Ward explained

WARD comprises the development of an intelligent clinical support system (WARD-CSS) that allows for 24/7 monitoring of oxygen saturation, respiratory rate, cardiac function and other parameters for patients. In WARD-COVID-19, patients suspected of COVID-19 will be monitored by this system, which includes a series of wireless sensors that transmit data to a server, and an artificial Intelligence (AI) that continuously monitors the incoming data and alerts staff to potentially critical situations. This represents a marked improvement over today’s intermittent manual observations that carry an inherent risk of patient deterioration between observations as well as contamination of staff and equipment. The WARD-COVID-19 project duration is set to eight months, and it will have clinical impact from day one.

Within a few days, infected patients can be hooked onto our monitoring system through wireless devices and thereby send important body condition data to staff via our computers. Within a few weeks, computer models can interpret data in real-time and warn doctors and nurses about critical symptoms if, for example, breathing deteriorates as a result of the coronavirus infection, mimicking the way clinicians interpret vital signs.
Physician, Eske K. Aasvang

Head of Research at the Department of Aneastesiology, Centre for Cancer and Organ Diseases at Rigshospitalet

Eske and his fellow project designer, chief physician Christian S. Meyhoff, head of research at the Department of Anaesthesia and Intensive Care at Bispebjerg and Frederiksberg Hospital, are preparing for the huge influx of patients expected to be admitted to the Danish hospitals with COVID-19 in the coming weeks.

Right now, the situation is getting critical and we are increasing the capacity at Danish intensive care units. But we can also see that the pressure on the general hospital wards is mounting, and that it is becoming increasingly difficult to monitor every patient with the current personnel and equipment. But we can use this new technology to make wards safer by monitoring and alerting staff about clinically deteriorating patients. This also means that the WARD system can be used to avoid many unnecessary hospital admissions, because with remote monitoring, we can wait until hospitalization becomes necessary. In this way, we can also protect both staff and patients from the spread of infection, and save a lot of protective equipment, which is currently in very short supply.
Chief physician, Christian S. Meyerhoff

Head of Research at the Department of Aneasthesia and Intensive Care at Bisbebjerg and Frederiksberg Hospital

The third designer of the WARD project, Professor Helge B.D. Sørensen at DTU’s Department of Health Technology, adds:

In the long run, the surveillance system’s artificial intelligence will be able to identify contexts and patterns in the development of disease. This is of course extremely relevant right now for corona-infected patients, where we still do not have a complete understanding of the patterns, but soon we can use the WARD system to predict the development of this disease by measuring oxygen saturation, breathing and several other parameters continuously. This will help doctors in both diagnostics and treatment later this year, when the second wave of the coronavirus is expected. At that point, the WARD system will be a huge asset with its ability to quickly and automatically interpret very big datasets.
Professor, Helge B.D. Sørensen

at DTU's Department of Health Technology