Blundur (“to doze” in Icelandic) is a non-invasive sensor to quantify respiratory effort during sleep, developed together with the Kempenhaeghe sleep institute.
This project in 60 seconds
A sleep apnea is a chronic disorder where the body repeatedly stops breathing during sleep. It’s estimated to affect 300.000 people in the Netherlands. The most reliable method to distinguish more complex types of apneas is by using respiratory effort. The current reference method to quantify respiratory effort is by means of inserting a catheter into the oesophagus. This method is uncomfortable for patients and sometimes poorly tolerated. Less invasive alternatives like thoracic belts, do not meet the requirements with regards to signal quality and reliability.
Kempenhaeghe, the largest sleep institute in the Netherlands, was working on a non-invasive sensor to quantify respiratory effort from the outside of the trachea, using a pressure transducer. Initial tests showed great results for short periods of time, but during the night the airtight pressure chamber would start leaking, rendering the sensor useless.
“In theory this method has a lot of potential, in practice it’s seriously hampered by the device-patient interface.”
In this project, I developed a patient-device interface that is both comfortable and stable and can be used reliably during a sleep study. I started the project by creating a data logger so I could test different iterations of the design during the night. Kempenhaeghe mentioned their interest in doing more measurements at home, therefore the last part of the project was spent on creating a home-monitoring solution.
To define the requirements for the final design, I discussed the goals for the project with the team at Kempenhaeghe and conducted a contextual inquiry to learn more about the context in which the sensor would be used.
Kempenhaeghe's goal was to have a well-functioning prototype to start early with clinical validation. For safety and to reduce the requirements for a clinical trial it was decided to keep electronics away from the body. In addition, it was decided to make use of rapid manufacturing services to easily be able to order prototypes.
An important insight from the contextual inquiry is that sensors are applied hours before the actual sleep study. Therefore it should be possible to connect and disconnect the sensor comfortably from the data acquisition system.
Based on these requirements I started a process of designing, testing and validating by collecting data about the sensor response.
On the left, you see Kempenhaeghe's original design for the patient device interface. Due to the protrusion, the contact area is small and constant counterpressure is given on the double-sided tape. My initial proposal was to create a flexible surface area as can be seen on the right. The flexible surface follows the skin, increasing the contact area and distributing forces better.
I decided to start my design process by creating a data collection circuit to quickly validate different iterations of the sensor system during the night. The circuit utilizes a low-pass filter, amplifies and converts the analog signal from bipolar to unipolar before saving the sensor signal with a timestamp to an SD card.
I started by experimenting with silicone rubber and scrap materials but wasn't getting an airtight seal. While doing research, I learned about tracheostoma patches used for laryngectomy patients. Laryngectomy patients breathe using an opening in their neck. The air filters and voice prostheses they use, have to be held airtight against the skin at the same location as Kempenhaeghe's proposed sensor.
I created a quick proof of concept to validate with the data collection system. While testing during the day I got good results for up to an hour. This approach seemed promising.
One of the first functioning prototypes made with a tracheostoma patch and some scrap materials.
Testing and iteration
After the promising initial results, I created a 3D printed connector. I aimed at improving reliability, with fewer connections that could break or get loose. While testing this connector over multiple nights, I got my first successful night without leaks. The problem, however, was that sometimes the tube would come loose during the night due to moving or stretching the body.
To improve comfort and decrease force on the connector, I created a strain relief by putting the connector on an angle. It creates a curve in the tube, which functions as a buffer while moving or stretching the body.
Attaching the tubing to the connector needed a lot of pressure, which was unpleasant on the body. I implemented a "Luer Lock", which is designed to create a secure and leak-free connection for gasses and fluids within a medical setting or laboratory. The connection is secured by twisting the connector, resulting in a secure, easy and more comfortable attachment of the sensor to the data acquisition system.
The new design was tested during the night with people from different genders and ages. The new design was very stable and delivered consistent results from night to night.
Some iterations made for the patient device interface.
The graph below shows how the sensor response changes for normal breathing on the left and obstructed breathing on the right. Obstructed breathing is simulated by breathing through a straw. It results in higher and wider peaks as it takes more effort and time to obtain enough oxygen.
After developing the sensor into a stable design that is easy to apply, the last part of the project was spent on developing a home-monitoring solution.
“We’ll start with comparing the different sensor technologies in a clinical setting, but the intention of these kinds of technologies is that they potentially could be applied in the setting of your home.”
Kempenhaeghe showed interest in home-monitoring, to gain long-term sleep data of the patient in their “natural environment.” For me, it was an opportunity to continue exploring product design and prototyping. My goal was to find a balance between a medical device and a consumer product that would fit within the home of the patient. Because the physician needs to be able to quickly explain and hand-over the device after a consultation, the focus was on making the device very simple and plug and play.
The prototype was created by stacking lasercutted parts together, filling and a lot of sanding to get a smooth surface. The top is created by layering magohany veneer and cutting/engraving it using a lasercutter.
The goal of the home-monitoring solution is to make it very easy to start a measurement. The device starts collecting data as soon as it's powered on. A small led on the back of the device will start pulsing to the rhythm of your breathing, to reassure that the sensor is properly applied to the skin. The LED is only visible when connecting the sensor and will not show at night to prevent distraction from sleep.
Below some of the product photo's shot at the end of the project:
With my designs and research data I went back to Kempenhaeghe to evaluate the project and discuss future opportunities. The team of Kempenhaeghe responded very positively.
“Since the solution is so simple and non-invasive, we are quickly able to start testing with patients. With consent of the patient, we are able to add the sensor to the polysomnography study quite easily.”
“We didn’t dare to dream of such a solution! I have high expectations!”
After finishing the project, Kempenhaeghe started testing the sensor with multiple patients with positive results. They are currently working on the scientific validation of the sensor. To be continued!
Dutch Design Week
Blundur was part of Mind the Step 2015, an exhibition about design and technology during the Dutch Design Week in the Klokgebouw.
I hope you enjoyed this project.
How about looking at another one below?