What can be measured with the device?
Dr. Kreuzer: The standard parameters include heart rate, body core temperature, oxygen saturation in the blood (SPO2) and respiratory rate. These are the basic parameters, from which further parameters such as stress level or fatigue level can be derived. All these parameters are based on heart rate variability (HRV). Other variables are then included in the medical set-up to use it as an early warning score system.
An example scenario where our device is used would be after an accident. When the patient is admitted to the clinic, various parameters are used to assess the patient's condition. Here the sensors can be used to check the patient's oxygen saturation or temperature.
How is the data managed and evaluated?
Dr. Kreuzer: In the field of sports, cosinuss° one can be directly linked as a sensor to your own Smartwatch or fitness application. In the clinical or medical field, the data can be pulled onto an app so that it can be accessed locally by the patient or user.
One goal for the future is that the data can also be stored on the sensor itself. This would allow the sensor to run autonomously and record data throughout the day, which would then be synchronized and downloaded with a mobile phone. This was planned as one of the next steps in the development, but now Corona is slowing us down a bit, as we are currently focusing entirely on solutions for Covid-19 patients.
To what extent is the sensor used for Covid-19 patients?
Dr. Kreuzer: Our sensor is now being used for Covid-19 patients who must stay at home because the clinics have to be kept clear for severe cases. For the patients this means that if you feel worse, you should admit yourself to the clinic. But here the question arises, what does worse mean? In Italy, there was a study that showed that older men particularly admitted themselves to the clinic too late and then arrived in such a desolate state that they could not be rescued. Our sensor supports the patients at home, because the data is transmitted to the health authority or clinic and with the help of an evaluation it is immediately visible if the condition of a patient deteriorates. The patient is then contacted and brought to the clinic before the condition becomes too bad.
How is artificial intelligence integrated into the device and the data management?
Dr. Kreuzer: We use AI for the development of algorithms. We stream the raw data to the server and then apply the algorithms using machine learning, and then re-apply the optimized algorithm to the sensor. At the same time, we run algorithms on the collected data to detect patterns. However, due to regulatory approvals in the medical field, this is all still very experimental.