Accurate, Wrist-Based Heart-Rate Monitoring—Six Months Faster

October 11, 2018

Andrew Burt By: Andrew Burt
Executive Business Manager, Industrial & Healthcare Business Unit, Maxim Integrated 


Concerned about a potential heart condition, the patient gets ready for her doctor’s appointment. But instead of traveling to the healthcare facility and waiting for her name to be called, the patient logs onto her home computer and signs in to a video conferencing app. Her doctor, meanwhile, has reviewed her history plus the vitals that her wearable device has collected over the past few weeks and is ready to provide her with some guidance.

Down the road, virtual doctor visits like this one could become the norm, with wearables playing a significant role. A recent study by The Scripps Research Institute found that a wearable ECG device could help detect undiagnosed atrial fibrillation. A randomized group of 2,659 participants were in the study, and 1,738 of them completed a period of active monitoring using the iRhythm Zio, a water-resistant ECG monitoring skin adhesive patch. According to the study, those who immediately began the monitoring process were significantly more likely to receive an atrial fibrillation diagnosis compared to those who began the monitoring four months later. In its September announcement of the Apple Watch Series 4, Apple notes that the device will have electrodes built into its digital crown that users can touch for 30 seconds to generate an ECG reading. The app has been cleared by the FDA, though the agency notes that the ECG data is for informational purposes only. This development demonstrates that the wrist is a valid area for taking ECG readings. Continuously collected health data, meantime, is gaining traction as a means for disease management and preventive care. For instance, also in September, Fitbit announced Fitbit Care, a connected health platform for health plans, employers, and health systems. Using data gathered by the company’s wearable devices, Fitbit Care would provide appropriate health coaching and virtual care to participants.

Many of us are now used to tracking our steps, calories burned, and our heart rate from fitness bands and smart watches. At the same time, however, we also tend to wonder how accurate these devices are. Device accuracy varies based on a variety of factors, from where the device is placed on the body (chest, wrist, or finger, for example) to differences in the optical properties of the skin. Wearable chest straps have proven to be an effective means for precise electrocardiogram (ECG) monitoring. However, a chest strap is not something that many of us would feel comfortable wearing daily. A wrist-based form factor is much more convenient and comfortable for continuous monitoring—but deriving accurate ECG data from this format has been challenging.

Health Sensor Platform 2.0 supports development of wrist-based applications that accurately measure ECG

As more of us rely on fitness trackers to monitor our vitals, we also need assurance of their accuracy.

What needs to happen to generate more accurate heart-rate data from wrist-based wearables? An ECG provides insight into the electrical activity generated within heart muscles. Heart rate is among the information that can be gathered when these differential biopotential signals are detected and amplified. Wet and dry electrodes attach to the skin to receive the ECG signals; the size and materials used in the electrodes can affect signal quality and levels detected. For a reliable signal, the electrode must have a good connection to the body. Dry electrodes present a more convenient format for wearers; however, with this approach, the ECG channel’s analog circuitry should have a very high input impedance to minimize the attenuation and small signals that stem from dry electrodes.

ECG signal quality is also impacted by motion artifacts that result from the wearer’s activities. A high common-mode rejection ratio (CMRR) of the analog front-end can minimize the effects of these motion artifacts. As with other wearables, it’s also important to minimize power consumption in order to extend battery life and keep the device cool. In addition, we cannot overlook the quality of the algorithms that turn the data gathered into actionable information.

Enabling a Healthier World Via Personalized Healthcare

In its latest foray into wearable healthcare technologies, Maxim has introduced a solution that overcomes the accuracy challenges for wrist-based devices. The Health Sensor Platform 2.0 (HSP 2.0), also known as MAXREFDES101#, is a next-generation, modular rapid prototyping, evaluation, and development solution for wearable applications that continuously measure heart rate, ECG, and body temperature. Available in a watch casing form factor, the platform gives you the flexibility to quickly create unique, highly accurate applications, saving up to six months in development time. By bringing together photoplethysmography (PPG, a process that measures the volumetric change of blood in tissue during the cardiac cycle), ECG, and body temperature data, collected on a continuous basis, the platform can enable designers to create applications that uncover a variety of meaningful health and well-being insights. Since the platform is open, you can also use it to evaluate your own algorithms. And because it is modular, the platform will quickly accommodate new sensors as Maxim introduces them to the market.

HSP 2.0 provides basic functionality out of the box and allows measurements to start immediately. Contacts on the back of the watch and the metal contact on top measure the small voltage for the ECG reading. Data can be stored on the platform for patient evaluation or streamed to a PC for high-level data extraction. Device users own the data collected, and they can tap into .csv files of the measured data for algorithm development or cloud-based data analysis. Because users own the data, this could open up opportunities for development of new, cloud-based “Big Data” and artificial intelligence (AI) algorithms that perform analytics, perhaps aggregating anonymous data points to uncover health trends. The platform, which has Arm® Mbed™ support, includes:

  • MAX32630 DARWIN ultra-low-power Arm® Cortex®-M4 microcontroller with floating-point unit (FPU) for wearables and internet of things (IoT) applications
  • MAX32664 ultra-low-power biometric sensor hub with embedded heart-rate algorithm
  • MAX20303 wearable power management solution featuring power-optimized voltage regulators and built-in battery management
  • MAX30205 human body temperature sensor with ±0.1 degree Celsius accuracy
  • MAX30001 ultra-low-power, single-channel integrated biopotential and bioimpedance analog front-end (AFE) solution for wearable applications, providing clinical-grade ECG readings
  • MAX86141 ultra-low-power optical pulse oximeter and heart-rate sensor for wearables

HSP 2.0 complements Maxim’s recently introduced MAX-HEALTH-BAND, an evaluation and development platform that streams raw data from sensors to output heart rate, heart-rate variability, step count, activity classification, and calorie consumption for health and fitness applications. The company also recently unveiled MAX-ECG-MONITOR, a wearable ECG and heart monitor evaluation and development platform, available in a wet electrode patch for clinical applications and a chest strap for fitness applications.

You can see HSP 2.0 in action and learn more about the platform at electronica 2018 in Munich, Germany, in November. We'll have demos and more information in Hall C4, Stand 440. Meantime, you can get a deeper dive into ECG measurements from our tutorial, "How to Measure Biopotential ECG Using a Chest Strap." While HSP 2.0 is targeted toward wrist-based applications, the insights in this tutorial on obtaining accurate ECG readings are relevant.