Remote Health Monitoring with FaceVitals

Summary

FaceVitals uses Transdermal Optical Imaging (TOI) to capture health data from a user's face.

Transdermal Optical Imaging (TOI) uses a digital camera to estimate vital signs by analyzing the way light travels through the skin and is re-emitted to extract information about blood flow, the blood vessels, and other structures beneath the surface of the skin. FaceVitals employs TOI to surface basic health vitals.

The process begins by using ambient light, which penetrates the skin and reaches the tissue below. Some of this light is re-emitted and captured by a camera. The color of that light contains information about both the blood and the skin. The information is separated based on the different colors of hemoglobin (in the red blood cells in the blood) and melanin (in the skin) using advanced machine learning algorithms. These algorithms process the information to create a map of the blood flow across relevant areas of the face, which is then linked together to form a video of the changes in blood flow over time. Utilizing this blood flow map, various vital signs can be estimated.

Blood pressure is the measured force that blood exerts against the walls of the arteries as it flows through them. As the heart pumps blood, the blood vessels in the face will expand and contract. TOI can measure blood pressure by analyzing how much the shape of the blood vessel changes with each expansion and contraction, estimating blood pressure.

‍

‍

The heart rate is the number of times the heart beats per minute. This beating of the heart causes changes in the diameter of blood vessels. By measuring the rate at which these blood vessels change diameter, the heart rate can be estimated.

TOI can also measure oxygen saturation. Oxygen saturation is the amount of oxygen in the red blood cells in the blood. The more oxygen in the blood, the more hemoglobin is saturated with oxygen. TOI can measure the concentration of hemoglobin in the blood through the unique way it absorbs certain colors light, which can give then be used to calculate an estimate of the oxygen saturation of the blood.

TOI is often used as a non-invasive alternative to arterial blood gas analysis (ABG) where invasive methods are not practical or desirable. By analyzing these changes, the data signals captured by TOI are then passed through patented and proprietary calculations and algorithms to estimate health vitals.

‍

Further information regarding TOI research can be found here: https://bit.ly/2zbDgFk

Summary

FaceVitals enables a quick and easy way to estimate vital signs and predict risks associated with chronic disease. Using the camera on your smartphone or other web-connected device, FaceVitals can predict estimated heart rate, respiration rate, and blood pressure in under 30 seconds. Our technology is supported by clinical studies in over 2,000 candidates to validate the accuracy of measurements. The data from the studies were used to create training models in order to predict and provide risk ratings for cardiovascular disease, stroke, and heart attack.

When a User accesses FaceVitals for the first time, they will be asked to input their personal details such as:

Gender at birth

Height

Weight

Age

As well as other information regarding their health such as:

Smoking history

Diabetes

Hypertension

Blood medications

This information is only required once.

The user will then follow an on-screen step guide on how to complete a successful scan. Directions will include how to position the face, removing intrusions such as heavy make-up, eyewear, baseball caps/hats, and ensuring the user has even lighting across the face.

At the bottom of the screen, a 1-to-5-star rating will indicate the quality of the scan being captured. Users are encouraged to reach 5 stars for a highly successful scan, however, the minimum acceptability criteria is 3 stars.

The scan itself takes a total of 30 seconds, and the results will appear on screen in the seconds following a scan’s completion. If the conditions are correct, then the scan will typically take less than a minute to complete. A user can experience longer completion times if they are needing to adjust for the minimum star rating.

Summary

While a user is conducting a FaceVitals, the below is what is happening from a technical perspective.

Step 1: Real-Time Scan

We first conduct a 3-second calibration allowing us to lock onto key areas of the face and to adjust the contrast of the image for a perfect reading.

Once calibrated, a 30-second scan begins measuring those key points on the face, securely sending the recorded measurements to the cloud.

Behind The Scenes

‍We use advanced signal processing to read the minute changes in color of the sub-dermal layer of the skin. This works on any skin type and tone.

Step 2: Scan Processing

After the scan process has been completed, the recorded measurements are then processed in the cloud into results.

Behind The Scenes

Using our cloud-based deep learning neural network, we process the recorded measurements and provide a result.

Summary

FaceVitals data is pre-processed on device for de-identification purposes.

Currently, data collected from a scan is pre-processed on the device. This is done for de-identification purposes. To remove the need of sending a User's image or video content to the cloud, the de-identification process converts the FaceVitals video to signals. Those signals then get sent to the cloud so results can be processed.

Categories of personal data

The types of personal data we collect depend on the specific Services we provide and your role. If you are a user of our FaceVitals product, a patient, a practitioner, or represent a healthcare provider who is one of our customers, we now store your personal data. In all other circumstances, we may collect some of the following information from you when you use our Services. The table below outlines the categories of personal data we may collect and the sources from which we obtain your personal data.

Legal basis for personal data processing

The table below delineates the legal foundation that permits us to process your personal data. While not all countries where we collect personal data adhere to the same legal standards, we uphold the highest possible safeguards to protect your data. Rest assured that whenever we process your personal data, we do so with a legitimate basis.

Summary

No content containing images or videos is sent or stored anywhere.

Upon completion of a FaceVitals, all content regarding the user’s image will be immediately deleted from the user’s device. This is to ensure privacy.

The User will be asked to grant permission for FaceVitals to access the device’s front camera, however, that is not a request to capture or store data.

Users do not have access to save or store any content regarding their images. The only time a User will see their image is during the scan itself when the camera is on, images are then immediately deleted after the results are returned.

Summary

The quality of a FaceVitals can often rely on the user's behavior and environment.

There is no set preference as to whether a User should hold the smartphone in their hand or place it on a steady surface during the scan. However, the smartphone must not be continuously moving around during the scan.

It is advised that the User find a steady surface, and have the phone face a clear uncluttered background, allowing the User to focus purely on lighting and face alignment within the on-screen circle.

Elements that could make a scan unsuccessful:

Heavily applied make-up, freshly applied sunscreen, or creams.

Accessories around the face like hats, scarves, or glasses.

Poor lighting or uneven lighting. TOI technology needs a human's whole face illuminated well and evenly. Users with half a shadow across their face may encounter unsuccessful results. A User may need to move closer to light or change settings if the lighting is not adequate.

Multiple faces on screen. FaceVitals was designed to capture the face of one person, not multiple. Any additional faces on the screen will confuse the technology and deliver inaccurate results.

While it’s not a requirement, Users can also use ring lights to successfully illuminate their face. During FaceVitals trial and validation studies, it was found that ring lights are the best method of achieving optimal illumination of the face.

Summary

FaceVitals was trained and validated on more than 2000 research participants.

To capture human data for training and validation of FaceVitals, 2,348 subjects predominantly aged 18+ were recruited over a course of two years, at the University of Toronto (Canada), and the physical examination center of the Affiliated Hospital of Hangzhou Normal University (China).

The use of human subjects in this study was approved by the institutional review committees at both institutions. Subjects participated in a series of events capturing the data needed to train the machine learning models, as well as participating in the use of FaceVitals technology for validation purposes.

To ensure accuracy in measurements, the computational models were also trained and validated against already well-established scientific instruments found in labs and clinics. An example of this was a proof-of-concept study involving the comparison of our technology against a traditional automated blood pressure monitor, to which accuracy was comparable.

FaceVitals, utilizes a comprehensive method that captures light reflected from a subject's face, subsequently analyzing it using advanced machine learning to measure changes in facial blood flow. The technology has been subjected to numerous research initiatives and peer-reviewed studies. These endeavors were geared towards ensuring the accuracy, reliability, and efficacy of the FaceVitals method. The results derived from FaceVitals have been corroborated through these scientific investigations, establishing its credibility.

For a more granular exploration of the validation processes and the corresponding results, the company is willing to share detailed research and study findings, contingent upon executing an appropriate Non-Disclosure Agreement (NDA).

Important Note on FaceVitals' Medical Status

It is paramount for potential users and administrators of FaceVitals to recognize that, as of now, this technology is not a medically approved device. However, it's worth noting that Transdermal Optical Imaging (TOI) is in the process of being submitted for FDA approval, reflecting the company's commitment to meeting rigorous medical standards. Consequently, when contemplating its application across various use cases, this non-medical approval

status should be prominently considered to guarantee its suitable and conscientious deployment. The technology's foremost objective is to furnish insights based on the data it captures rather than serving as a definitive medical diagnostic tool.

For stakeholders interested in the scientific foundation of TOI, the company stands ready to provide access to a wealth of research and peer-reviewed studies that underscore the technology's capabilities and potential. However, to ensure the confidentiality and integrity of this information, it will be shared only once an appropriate Non-Disclosure Agreement (NDA) has been duly executed. The company remains dedicated to transparency and collaboration, working diligently to drive advancements in the field.

Summary

A number of data outputs become available after doing a FaceVitals.

Accessible Data

A simple data output (A JSON compatible object) is generated from each successfully completed scan. This provides flexibility to interpret and present it where it makes the most sense for your users.

Your Data, Your Storage

FaceVitals are processed in the cloud (AWS) however, results are stored to allow the healthcare practitioner access to the results.

Data Privacy

Due to the way we record measurements, no photos or videos will ever leave the device.

FaceVitals requires the following biometric inputs from users to help produce the scan results.

User input required:

Height

Weight

Sex

Age

Do you smoke?

Do you have hypertension?

Are you on blood pressure medication?

Are you diabetic?

Scan results calculated:

Heart Rate

Estimated Blood Pressure (Systolic & Diastolic)

Irregular Heartbeats

Cardiac Workload

Breathing Rate

Heart Rate Variability

Cardiovascular Disease Risk

Stroke Risk

Heart Attack Risk

‍

Summary

Hypotension, or low blood pressure, is a medical condition in which the force of blood against the walls of the arteries is consistently lower than normal.

In some cases, it can lead to symptoms such as dizziness, fainting, and fatigue.

Hypotension can be caused by various factors such as dehydration, sudden blood loss, heart conditions, and certain medications. It is important to work with a healthcare provider to determine the cause and develop an appropriate treatment plan.

Summary

Heart Attack Risk is an individual’s likelihood of experiencing their first heart attack within the next 10 years, expressed as a percentage.

The risk score considers a subject’s biological sex at birth, age, BMI, systolic BP, and antihypertensive medication status as well as their smoking status and diabetes status when available.

A heart attack occurs when a coronary artery, which supplies blood to your heart, becomes blocked. This stops the blood flow and reduces the amount of oxygen that gets to your heart muscle. This causes the most common heart attack symptoms, which are chest discomfort or pain and can spread to your arms, neck, jaw or back. Unlike angina this discomfort often last for longer than 10 minutes.

Summary

Hypertension Risk corresponds to the percentage of people with the subject’s risk profile who are diagnosed with hypertension (high blood pressure) by their doctor.

The risk profile is based on the subject’s facial blood flow (using Transdermal Optical Imaging) and their profile information (biological sex at birth, age, height, weight).

Moment-to-moment fluctuations in the subject's physiology mean that their risk estimates could vary to some degree from one measurement to the next. The best estimate of the subject's overall risk is therefore obtained by averaging several measurements throughout the day and across several days to adequately account for this physiological variation.

Summary

Cardiovascular disease (CVD) is a class of diseases that involve the heart or blood vessels. CVD includes coronary artery diseases (CAD) such as angina and myocardial infarction (commonly known as a heart attack).

Other CVDs include stroke, heart failure, hypertensive heart disease, rheumatic heart disease, cardiomyopathy, abnormal heart rhythms, congenital heart disease, valvular heart disease, carditis, aortic aneurysms, peripheral artery disease, thromboembolic disease, and venous thrombosis.

About the measurement

Cardiovascular Disease Risk is the likelihood (expressed in percentage) for you to experience a cardiovascular disease (heart attack, stroke, or peripheral artery disease) within the next 10 years. This risk score is based upon the Framingham Method and derived from an algorithm based on data from prospective studies that followed up participants in terms of their cardiovascular health for over ten years.

Summary

Blood Pressure consists of Systolic and Diastolic readings and is usually written as two numbers Systolic / Diastolic e.g. 140/90 mmHg

Systolic blood pressure (the first number) – indicates how much pressure your blood is exerting against your artery walls when the heart beats.

Diastolic blood pressure (the second number) – indicates how much pressure your blood is exerting against your artery walls while the heart is resting between beats.

Typically, more attention is given to systolic blood pressure as a major risk factor for CVD for people over 50.

Either an elevated systolic or an elevated diastolic blood pressure reading may be used to make a diagnosis of high blood pressure.

The risk of death from ischemic heart disease and stroke doubles with every 20 mm Hg systolic or 10 mm Hg diastolic increase among people from age 40 to 89.

Summary

Breathing rate or Respiratory Rate is the number of breaths you take per minute.

The normal respiration rate for an adult at rest is 12 to 20 breaths per minute.

A respiration rate under 12 or over 25 breaths per minute while resting is considered abnormal.

Categories

Abnormally Low

Breathing rate per minute: 0.00 to < 12.00

Normal

Breathing rate per minute: 12.00 to < 25.00

Abnormally High

Breathing rate per minute: 25.00 to < 42.00

Summary

Cardiac workload is a measure of stress put on the heart muscle.

Cardiac workload, more specifically known as myocardial workload, is a measure of stress put on the heart muscle. It refers to the amount of work performed by the heart in pumping blood through the body.

Cardiac Workload can be calculated using the formula: Heart Rate (bpm) x Systolic Blood Pressure(mmHg). This is shown in dB (decibels), measuring signal to noise ratio.

It is determined by the amount of blood pumped, the pressure at which it is pumped, and the heart rate. An increase in cardiac workload can be caused by physical activity, disease, or other factors and can have an impact on heart health.

Summary

Heart Rate Variability is a measure of the variation in time between consecutive heartbeats.

Heart Rate Variability (HRV) is a measure of the variation in time between consecutive heartbeats. The standard Deviation of normal-to-normal (SDNN) is a measure of HRV. SDNN is the standard deviation of the time intervals between each heartbeat, expressed in milliseconds (ms).

It is an indicator of the body's ability to respond to physical and psychological stress and reflects the autonomic nervous system (ANS), which regulates important bodily functions such as heart rate, blood pressure, and breathing.

A low HRV can indicate a reduced ability of the ANS to respond to stress and may be associated with an increased risk of health problems such as cardiovascular disease and depression.

Incorporating activities that promote HRV, such as exercise, deep breathing, and mindfulness practices, into one's daily routine may help improve ANS function and overall health.

Note: Our standard suggested range for HRV is 0-80ms, however any figure >80ms can be considered Very High and in the Low Risk zone.

Summary

Stroke Risk is an individual’s likelihood of experiencing their first stroke within the next 10 years, expressed as a percentage.

The risk score considers a subject’s biological sex at birth, age, BMI, systolic BP, and antihypertensive medication status as well as their smoking status and diabetes status when available.

Stroke can lead to temporary or permanent disability, including paralysis, speech problems, and memory loss. Some common risk factors for stroke include high blood pressure, smoking, obesity, physical inactivity, and a history of heart disease.

Preventing stroke can involve making lifestyle changes, such as eating a healthy diet, exercising regularly, quitting smoking, and controlling risk factors such as high blood pressure. Early recognition of the signs of stroke and prompt treatment can greatly improve outcomes and reduce the risk of serious complications.

Summary

FaceVitals are compatible with iOS and Android platforms.

FaceVitals will work on iOS operating systems 12.1 or above, and Android operating systems running version 8 or above (64-bit and OpenGL 3.1 supported).

All types of cameras on the devices within the range above, will be able to conduct scans.

FaceVitals is functional for people aged 13+

FaceVitals is effective for the ages of 13 and older. However, results for 13-17-year-olds may not be as consistent as people who are 18+.

When recruiting individuals for research projects, the vast majority of subjects will be 18+ giving us far greater data to train the machine learning models.

Yes, FaceVitals will work for all skin tones.

The most important thing to note when conducting a FaceVitals is to achieve even lighting across the face. Given that TOI needs to capture re-emitted light from blood flow, the face must be illuminated well and evenly.

It is worth noting that a User with darker skin, may require more light across the face if having trouble conducting a successful scan.

While it’s not a requirement, Users can also use ring lights to successfully illuminate their face. During FaceVitals trial and validation studies, it was found that ring lights are the best method of achieving optimal illumination of the face.

Summary

FaceVitals is currently not certified as a medical device. But is going through the process of FDA approval.

FaceVItals technology is not currently certified as a medical device, however, is currently under the process of FDA (Food & Drug Administration) approval in the United States. Upvio will release further information regarding the FDA status over time, with a desired goal of approval in 2023/24.

FaceVitals is also compliant with several global data regulations (APA, CCPA, GDPR (EU and UK), LGPD, PDPL, and PIPL).

Summary

FaceVitals is a non-intrusive and contactless form of capturing vital signs and health risk indicators.

This technology does not require Users to purchase or onboard new hardware and it is easily accessible on a smartphone.

FaceVitals is a building block for scalable healthcare solutions, giving healthcare organizations an accurate and reliable remote monitoring ability via a device already widely circulated (the smartphone).

FaceVitals carries the potential to be a cost-reducing tool regarding hardware within organizations. With traditional hardware models requiring regular maintenance and upgrades. Adding to this, as organizations or patient loads expand, so do equipment needs; FaceVitals is ultimately scalable in a cost-effective way as it is delivered on a smartphone.

FaceVitals’ at-home or pre-appointment ability also increases the amount of time physicians or organizations can spend working with people on positive health outcomes.

Summary

Yes, an internet connection is required while conducting scans.

Internet is required at all times during the scan process. This is due to licensing requirements regarding ScanControl.

If there is no internet access on the device, scans will not work.

It is important to note that while the quality of the camera may differ from model to model, it will not have any significant impact on the accuracy of scans.

Summary

All measurements suffer from a natural deviation because of the nature of measuring the human body.

Two factors could influence differing results between scans. The first is that the human body is constantly changing. The second involves environmental factors including sitting vs standing, posture, lighting and the light direction, the distance of a face from the camera, face angle, Internet connection, and beverage and/or food intake.

In general measurements of any form (digital or physical) will not be consistent between captures. All measurements suffer from a natural deviation because of the nature of measuring the human body, for example, if a person repeatably measures themself with a tape measure they don’t get the same measurement every time due to breathing.

The Upvio team does not foresee any safety concerns surrounding people who are pregnant using FaceVitals. Nor will it impede results.