Blood sugar, pulse, blood pressure – doctors record and use all the information they can get. Thanks to wearables on the body, this information becomes more complete and can also be recorded at home. This primarily benefits those affected.
In pregnancy care, it is normal procedure to check the child’s heartbeat and other vital parameters, usually once a month. Towards the end of the pregnancy and during birth, the cardiac contraction recorder, the so-called CTG, is used. The technology is proven – but it has disadvantages, says Anda Radan, senior physician at the women’s clinic at the Inselspital Bern: “The size of the electrodes has not changed since the 1960s,” says the doctor.
Long-term goal: monitoring at home
With large sensors on her stomach, the patient has to lie on her back; as quiet as possible so that the recording device can receive and register the signals. Anda Radan believes that this cumbersome measuring method is no longer appropriate. Since 2019 she has been working on developing a comfortable pregnancy belt that works with so-called dry electrodes.
The long-term goal is monitoring at home. This would benefit many women, especially high-risk pregnant women, says Anda Radan: “If the mother becomes ill or if the child is not developing well, people today often decide to go to hospital. This could be avoided in the future through close home monitoring.” The safety of the hospital has its price: lying down for a long time is neither good for the mother’s health nor for her psyche – and not for the fetus either.
The pregnancy belt was developed by engineers from the CSEM technology innovation center in Neuchâtel. A department of the CSEM recently moved into the Bern Island campus. With the university and the canton of Bern behind it, the InselspitalBern wants to promote medical technology innovations itself.
Use your voice as a warning system
Christoph Stettler is clinic director and professor of diabetology. His vision is a warning system for people with diabetes that works via voice.
The risk of hypoglycemia
One problem with diabetes is the strongly fluctuating blood sugar: “In order to avoid hypoglycemia, those affected have to measure their blood sugar again and again,” explains diabetologist Chrsitoph Stettler from the Inselspital Bern.
Sensors that are permanently attached to the upper arm would have made measuring easier, “but they measure in the wrong place, namely in the subcutaneous fatty tissue.” As a result, the measured value always lags slightly behind the actual blood sugar value.
This “little lagging behind” can be fatal because blood sugar sometimes drops rapidly and those affected are no longer able to react in time.
As a doctor, Christoph Stettler notices when his patients have low blood sugar levels by looking at their voice: “Since my own father is diabetic, I have been familiar with such voice changes since I was a child.”
So he came up with the idea of using the voice as a warning system and teamed up with the CSEM.
Recognize the difference – AI can do it
In the laboratory, the researchers broke down the voices of diabetics – according to the pitch of the tone, rhythm, precision of pronunciation, etc. The test subjects had to recite certain sequences of sounds or sections of text twice: in the normal state and in the hypoglycemic state. “Thanks to machine learning and artificial intelligence “We were able to notice differences,” says Christoph Stettler.
These are very clear for some test subjects, but barely audible for others. However, the algorithm always registers the differences – at least in the pilot study. In the next step, the research group wants to test the whole thing in normal everyday life, with all its noise.
And when all these hurdles have been overcome? Then, says Christoph Stettler, it would be entirely conceivable that the smartphone could one day warn those affected: “Hey, you are hypoglycemic! Sit down and eat something sweet!”