The brave new world of contactless biosensing

Most wearables today, such as your smartwatch, collect a variety of health data seamlessly. But to do so, the biosensors must contact some part of your body, most commonly your skin, to collect health data. A new paradigm is starting to emerge in consumer health — contactless biosensing — which acquires physiological data without making bodily contact. 

Broadly speaking, there are two main approaches to contactless biosensing: 

1. Optical method

In this approach, developers exploit the visible-light range of the electromagnetic spectrum. Incident light on the skin results in both specular and diffuse reflection, and pulsatile information is extracted from the diffuse reflection and analyzed to derive metrics about heart rate, blood pressure, and other vital signs (Figure 1). This method is also known as remote photoplethysmography (rPPG), given that, just like standard PPG sensors in on-body wearables, data are extracted through volumetric changes in blood vessels: constriction and dilation.

Binah.ai is a leader in this space and develops a software platform leveraging rPPG to gauge vital signs using a smartphone or tablet camera as a light source, measuring the intensity of light reflection off the skin of a person’s face (cheek area). The software is deployed as a software development kit, which consumers can download and use on their personal devices. The platform can capture metrics like blood pressure, heart rate, heart rate variability, respiration, and O₂ saturation. Binah.ai has developed a strategic partnership with CareOS to embed its contactless biosensing technology into smart mirrors for home use. Binah.ai also has strategic partners in insurance, with companies like Sompo Himawari, Generali Group, and Momentum, to allow insured clients to monitor their vital signs and get recommendations toward better health. This is potentially a win-win as insurance companies can encourage healthy habits and reduce healthcare payouts, and consumers can stay healthier and pay lower premiums. While demonstrations for small sets of users appear promising, little information is available across wider groups. NuraLogix is another emerging startup that uses an optical approach for contactless biosensing. It has conducted trials and claims to have data from over 40,000 patients. It recently launched its first hardware product, Anura MagicMirror, which is a tabletop smart mirror for consumer use at home.

Contactless biosensing solutions must compete with invasive and noninvasive options, so precision and accuracy in measurements will be critical to success. A good metric to consider is blood pressure. The standard deviation for systolic blood pressure for home use should be within 12 mmHg (the official blood pressure metric) and within 8 mmHg for diastolic blood pressure, based on a universal standard for validating the devices. For clinical use, it should be 6 mmHg or lower. Binah.ai (11 mmHg) and NuraLogix (8 mmHg) report error rates acceptable for consumer home use but not yet at an acceptable range for clinical use. This accuracy is good enough for general wellness monitoring, but consumers should seek clinical evaluation for healthcare emergencies. Hence, the technology has made some progress toward general consumer health monitoring, although the findings are limited to small subsets of users.

2. Radio waves

Developers of this solution also exploit the electromagnetic spectrum but utilize the longer wavelength of radio waves to acquire physiological data. Radio waves’ lower frequencies can’t penetrate the skin as deeply as light can to extract vital physiological information. The benefit of using radio frequencies is that they have a lower energy per photon and so use less energy and see less scattering from the atmosphere, allowing for stronger incoming signals.

Neteera Technologies develops a contactless, sub-terrahertz, radar-on-a-chip device that can measure vital signs (heart rate, heart rate variability, blood pressure, and respiratory rate) based on micro-skin displacements caused by the heartbeat and respiration, up to a distance of 5 feet away. It claims to have an accuracy comparable to the Mindray ePM 10M patient monitor, a vital-sign monitor used in hospitals. Neteera used the Mindray monitor as a reference device to show that its solution is “substantially equivalent” when submitting data to the U.S. FDA and subsequently received FDA clearance. The company is focused on the healthcare market, although it is interested in exploring consumer health opportunities. Neteera seems to be the only startup with a commercial product ready for trials in the radio wave category (technology readiness level 7). Academic institutions like The Ohio State University, University of Minnesota, and Jiangnan University are actively looking into developing this technology further.

In summary, contactless biosensing is a promising approach for two main reasons:

• It removes issues surrounding reduced retention rates: Studies have shown that wearables typically see a rapid decline in retention rate. One study of over 6,000 respondents from the U.S. found that 50% of users stopped wearing their device after six months. This is less of an issue with contactless biosensing as most users either have their smartphones with them or can place the device in their homes and not have to worry about putting on a separate wearable device.
• It can monitor data from multiple users without any restrictions: Unlike with wearables, it’s possible to monitor the data from more than one user (although not simultaneously). A parent, for example, could easily turn a smartphone loaded with Binah.ai’s app or an at-home device from Donisi toward his partner or dependent to check their vital signs as opposed to removing and reattaching a wearable, which may not fit individuals in different age groups.

That said, the current landscape of available players in contactless biosensing is sparse, but companies are beginning to penetrate the market. The limited availability and use-cases for consumers aren’t surprising, as available consumer data sets are limited to certain populations, and more work must be done to improve accuracy and precision. Data privacy concerns may slow penetration as well, especially for solutions that rely on transmitting short video clips.