Sensor-based solutions for vital signs measurements

Article By : Dr Chris Elliott FREng

We’ve come a long way from fuel and oil pressure gauges in vehicles, and now every feature is instrumented and logged, so we can diagnose faults after they occur but also anticipate them...

Our cars are full of sensors – we’ve come a long way from the fuel gauge and oil pressure and now every feature is instrumented and logged, so we can not only diagnose faults after they occur but also anticipate them. Increasingly we do the same with our homes, with smarthomes that adjust to the weather and optimise their use of resources. But we’ve been slow to do the same for our bodies.

The first generation personal sensors for health were little more than toys. They measured what was easy to sense, not what is useful to know. People are working hard to find medically-significant markers in the data that they collect but that is putting the cart before the horse – collecting data then finding a use rather than saying what we need then measuring it. Also, there is a good reason why FDA and the European CE Mark set such a high threshold for a medical device. History is full of snake-oil, and doctors are rightly sceptical of a mountain of data from a device of unknown provenance.

Figure 1: Author’s ECG trace from Kardia. (Source: By author from AliveCore’s Kardia mobile app)

The change came first with Alivecor (now called Kardia) who developed a one-lead ECG that links to a mobile phone. I made this record on my Kardia while writing this article – and before all the cardiologists call me, I know that I have left bundle branch block. I’m relaxed because I can check it regularly and confirm that it has not changed.

We are now starting to see other true medical devices that exploit sensors and the capacity of our phones to analyse their data. HealthyIO has a very elegant product that analyses the colour of urine-test dipsticks using the camera in a smartphone. Its brilliant trick is to provide a reference colour chart which allows the camera and lighting to be calibrated every time. The result is a cleared medical device that matches the accuracy of an expensive laboratory colorimeter but which is already in your pocket.

Figure 2: HealthyIO smartphone-based colorimeter system. (Source: Healthy.IO)

The next generation will be medical-grade sensors that are built into the phone, free to the user like the camera or GPS. My company has developed the V-Sensor, which measures blood pressure, temperature, blood oxygen, respiration rate and pulse rate in a device the size of a peanut and costing a few dollars. Bluetooth enabled and designed to work with our e-Checkup app, the V-Sensor is a Class II (IIa in Europe) medical device. The V-Sensor is designed to be built into a phone when the phone is made but, because of the way that it interfaces, the phone does not itself become a medical device.

Figure 3: Built into a smartphone, the LMD V-Sensor works with the LMD e-Checkup app to deliver vital signs measurements like other smartphone sensors such as accelerometers support location apps. (Source: LMD)

As well as being uniquely small and affordable, it has a novel and patented way of measuring absolute blood pressure. High blood pressure is a killer; on average one person dies every 4 seconds because their blood pressure is elevated. It can be easily treated if you know that you have it. That’s why absolute measurement is so important – many devices try to estimate the change in your blood pressure from the way that the pulse travels along the artery but they all need to be calibrated often with a cuff. If you’ve a cuff, you’ve already found that you have high blood pressure and don’t need the device. Our approach is different, because we emulate exactly what the cuff has been doing for over 100 years.

Like a cuff, we measure the pressure applied to the artery that just balances the pressure in the artery, with the twist that we measure on the fingertip and by asking you to press harder or softer to create the necessary pressure. This approach needs a great deal of computing power, both to communicate with the user to collect the data and to analyse the data to find the correct systolic and diastolic blood pressures, but a 2 GHz processor in the phone can deliver it. But it is worth the effort – the cuff has survived for all its inconvenience because it is accurate and it works, as does our solution.

We have engineered the device for mass production, including automated test and characterisation of the sensors, and have established manufacturing capability. As a medical device, this has to go through the full process to be awarded a CE Mark and for clearance by FDA, NMPA in China and all the other regulators and we are building the information dossiers for this.

Whether in a phone or in a wearable or standalone device that links to a phone or computer by Bluetooth, this provides the five basic vital signs for everyone. But it does much more. Our ASIC includes a one-lead ECG, like the Kardia, that measures between the fingertip and the fingers of the other hand. From our measured data, we can find the stiffness of the arteries and the Cardiac Output – the amount of blood pumped on each beat. By exploiting other sensors already present in the phone, such as the accelerometer, we can detect the timing of the heart valves and the dynamics of the heart movements. By tracking the measurements over time, we expect to be able to detect before they become critical many conditions, from sudden cardiac arrest to COVID-19.

Future versions of the V-Sensor will have other capabilities, some of which have already been patented ready for development. And, much as we would like it, we’re not the only people in the market. The race is one to turn every phone and every wearable into a true medical device, cleared by the regulators and measuring vital signs that underpin diagnosis and monitoring.

— Chris Elliott is one of the founders of Leman Micro Devices SA, based in Lausanne Switzerland. He is a Fellow of the UK’s Royal Academy of Engineering and also a regulatory barrister.

Leave a comment