Cleaning up the Signal to Noise in Cardiac Monitoring
By Matt Phillion
There’s no argument that cardiovascular disease is a massive issue both in the U.S. and across the world. A recent study found that roughly 35% of U.S. adults over 20 received care for cardiovascular risk factors, and costs related to cardiovascular care are expected to triple to nearly $1.3 trillion by 2050.
Meanwhile, cardiac monitoring has become more accessible than ever, as wearable devices capable of monitoring pulse and blood pressure join more advanced technology that can be used to monitor a patient’s cardiac health. The challenge here is twofold: first, the amount of data wearable technology generates is vast and overwhelming; and second, noisy signals from devices such as ECGs can result in unusable information, misinterpreted data, and lost time and effort cleaning up the signal and digging out the most relevant and accurate information.
With decades of deep R&D in ECG analysis and front-end design, B-Secur is a leading expert and innovator in noise and artifact detection and reduction to deliver medical accuracy and efficiency by offering better signal clarity and a reduction in ECG signal noise to provide actionable insights.
“We started off from the cybersecurity side, using ECG to form a new biometric modality as something more secure than fingerprints or facial recognition,” says Alan Foreman, CEO of B-Secur. “But within a year the data we were generating from human beings provided much more valuable insights into well-being and health.”
Nearly 18 million people die of cardiovascular disease each year, the largest cause of death in the world, with someone dying from cardiovascular disease every 34 seconds—and 80% of those cases are preventable, Foreman points out.
“With the cost to the U.S. economic system about to triple, it’s a big socioeconomic problem, and the way it’s dealt with is largely reactive,” says Foreman. “You wait until you have a problem, and then you’re hooked up to an ECG.”
And even with that reactive system, resources are scarce, Foreman notes, which is why he and his team are looking for a way to make ECG something that can be recorded anywhere—on your watch, in clothing, or other use cases—to help reach the greatest number of potential patients possible.
“It’s a way to start to self-assess what an issue is before it becomes a problem. Maybe it’s symptomatic stress or other sources that can lead to cardiovascular issues, but it’s a way to address them at a younger age and throughout your life,” says Foreman.
Better than good enough
Getting a better signal, leading to better data overall, is an important step to moving beyond just good enough, notes Tyeler Dean, VP of Medical with B-Secur.
“For the amount of repeat testing caused by noise—artefacts, muscle movement, unclean tracing—a lot of work is being done but nobody’s cracked the code,” says Dean. “We’re trying to drive an outcome that reduces noise in the data and leads to faster and more accurate diagnosis so patients don’t get trapped in a spin cycle of repeated testing.”
An example: a patient might be sent home with a Holter monitor for one, three, or seven days for an assessment. That data returns to the provider or hospital where the electrophysiologist or cardiologist will have to sift through the data before they can start to diagnose the patient. They look for what might be a real incident or event versus what is just noise recorded in an everyday environment.
“Holter monitors are very good but face an electrical signal problem. In many cases, half the data is redundant before the diagnosis can even start,” says Foreman.
“The benefit here is that providers and technicians are currently looking through a deluge of data. Even your Apple watch is creating a lot of information. How do you know what is actionable?” says Dean. “It’s knowing when and where to look.”
To filter the data, B-Secur has spent years training its algorithms to identify what clinicians need to see based on huge volumes of information.
“It comes back to the signal processing piece,” says Dean. “There’s a few different pieces to our software library that benefit the clinician but there’s no difference in the patient’s experience. It’s just improving the gain on the output.”
Dean also points out the software is agnostic to the hardware used, whether it’s a patch, a wearable, an implant, or other option. Dean describes the analysis as a red/yellow/green style of assessment—if it’s green the signal is good and incredibly accurate, and if it’s not it’s an area to focus on that needs more help.
“It’s a bit like comparing birdshot versus a bullet” trying to hit a target, Dean explains.
“Processing the patient shouldn’t feel any different except that you may get a quicker diagnosis,” says Foreman. “In the future, it may take an electrophysiologist half as much time to go through a data gathering exercise. This means these scarce analytical resources can double their output, in theory, and that has a huge impact on efficiencies within a healthcare system and the quality of patient care.”
Top of mind across all healthcare is that scarcity of resources—the industry simply doesn’t have enough trained professionals to meet the needs of its growing patient population.
“We know that in too many incidents of data collection, you’ve got to go back to the patient for repeating a Holter, patch, or mobile cardiac telemetry (MCT). It’s frustrating but also it has a real impact if you’re really unwell, and time is critical,” says Foreman. “Speed and efficiency in this space is something hospitals are looking for, and partnering with technology companies that understand the value.”
In terms of burnout, Dean also notes that not needing to re-test and shortening the time of analysis can lessen mental fatigue on staff and the patients they care for.
“This fatigue can impact the provider, caregiver, and patient. The mental anguish of not having a definitive answer,” he says. “There’s both an economic and mental cost associated with it.”
Cardiac health is the largest element of cost to the health system, Foreman notes, and is expected to go from 2.8% of the U.S. GDP to 4.7% in the next 25 years according to the study referenced above.
“There’s never been a better time to focus on this, because those numbers are not sustainable,” he says. “And patients are going to suffer for it. We’ve got to transform in a way that really changes the dial. It’s a way to take the hospital into the home.”
They are seeing a convergence in this space where not only healthcare systems and providers are looking at methods for improving the technology, or those tech companies familiar in the space, but big players outside of healthcare are also seeing an opportunity to collaborate with researchers to improve patient care.
“We’re at a crossroads of research and companies collaborating to reduce burnout and increase efficiency,” says Dean. “The organizations we’re working with are seeing between a 27[%] and 48% improvement on how much they get through from an accuracy standpoint, and there’s a domino effect that follows that.”
“We think we’ve brought a different set of eyes to the challenge and we hope what we’re doing can shed a light on the problem of electronic noise—and if that can be resolved, it opens up so many opportunities,” says Foreman.
Foreman says one of the goals is to maximize how powerful an impact continuous monitoring can have on patients overall.
“We set off on this mission to have ECGs everywhere. Wherever you look in a developed society, you can have it in implantables, patches, and smartwear devices, and that means you’re not just treating people after they get sick,” he says. “How far can you get in reaching 8 billion people in a way that can transform health systems everywhere?”
“We want to empower the users and patients to have an awareness of how their own data can intersect at different points in their lives,” says Dean.
Matt Phillion is a freelance writer covering healthcare, cybersecurity, and more. He can be reached at matthew.phillion@gmail.com.