How Facial Blood Flow Patterns Can Usher a New Era of Health Vital Monitoring

Those of us old enough to recall the 1960s TV series “Star Trek” were fascinated by Dr. McCoy when he would simply point a scanner at his patients and analyze their condition. What was once science fiction has now become a reality and the potential future state of medical device technology for the coming decades. This future state has been made possible since the advent of “Transdermal Optical Imaging” or “TOI^TM”. This novel form of remote photoplethysmography was the first globally to demonstrate the feasibility and accuracy of measuring blood pressure by analysis of facial blood flow captured from cameras of smartphones. ^1.

This revolutionary approach to contactless BP and vital signs measurement opened the door to creating new standards for device accuracy. The only existing standards for BP device accuracy of measurement would be the ANSI/AAMI/ISO81060-2:2018 which references cuff-based technology TOI^TM technology demonstrated the accuracy of blood pressure measurement according to these standards and these findings were published in 2020.^2.    

Work is now underway within IEEE SA 1708 standards for cuffless blood pressure devices working group to develop the best methodology for determining the accuracy of devices measuring blood pressure that do not use inflatable cuffs. Industry leaders, IEEE SA members, and the FDA are all working together to create accuracy standards for this innovative technology. Once cleared by the FDA, these medical devices will be Class 2 software as medical device technology. Applications of these technologies within current clinical workflows once FDA-cleared will change the face of medical monitoring and non-communicable disease management as we know it. The significant advantage to TOI^TM -based technology for BP management is freeing patients from the required purchase of devices other than their smartphones to measure their blood pressure.  Smartphone technology is available to 70% of people within the lowest 20% income nations.^3.

This technology, however, goes well beyond simply capturing vital signs. TOI^TM can capture other blood biomarker risks, and disease risk profiles can be assessed in a 30-second video scan. Foundational research published in 2021 advanced this technology even further in its ability to capture patient parameters that included multiple biomarker risks and body fat composition analysis.^4.

To understand how this is possible it’s important to understand the principles behind this technology.

Every beat of the heart produces a pulse wave, the rush of blood out into our extremities. Scientists have been studying the pulse wave for hundreds of years. Etienne- Jules Marey was the first physiologist to capture the shape of the pulse wave using a device called plysmograph. A complex array of weights and pulleys connected to a pencil would draw out the shape of the pulse, the rise and fall of blood volume within the artery at the wrist. How ironic this early researcher of the 1850s, also was an innovator in capturing moving picture imagery. That his two foundational technologies would collide one day into the miracle of contactless patient measurements, would have no doubt created joy for Etienne- Jules Marey.

Moving ahead 140+ years with the advent of biosensors and light technology, it became possible to capture the pulse wave by shining a direct light source through the skin. This technique is called Photoplethysmography, it’s foundationally like Etienne-Jules early methodology, only using light and sensors instead of weight and pulleys applied to the artery at the wrist. Most of the wearable technology today uses this photoplethysmography method to capture the vital signs of BP and HR. 

TOI^TM is yet one more step forward in the pulse wave analysis. Rather than using a direct light source shining into the skin, any smart device’s camera can capture the pulse waveform using reflected light. Natural light coming into the face is absorbed and the amount reflected will change depending on the change in blood flow related to the cardiac cycle, that rise and fall of blood pulsing through the face can be seen on video imaging of reflected light across the red, blue, and green spectrum. 

The facial blood flow patterns are incredibly complex and vary within regions of the face. The pulse wave patterns in the cheeks and nose behave differently than the pulse wave patterns of the chin and forehead, each area unique in its innervation and response to either the sympathetic (fight or flight response) or parasympathetic (rest and digest response) nervous system. The signals captured in over 23 regions of the face during video analysis can very accurately be translated into the pulse waveform like those captured during other forms of plethysmography.

How does one translate this data of waveform, thousands of features captured, and blood flow analysis, into accurate predictions of vital signs and disease risk profiles? 

Enter the magic of AI Machine Learning. TOI^TM also uses advanced AI ML models to make sense of all the data and features captured during the video analysis of facial blood flow. To build these models, however, requires access to tens of thousands of people across a diverse population set of which there is access to their “ground truth.” The ground truth would be parameters like BP measured accurately by auscultation method at the time of the scan, and multiple blood analyses for biomarkers like HbA1c, cholesterol, fasting blood sugar, hemoglobin, cardiovascular risks, and other disease states. Capturing the pulse wave beat to beat means that algorithms can be applied to calculate heart rate variability, a key component to mental stress measurements that show the balance between the sympathetic and parasympathetic nervous systems. The AI ML algorithms can be “trained” to recognize the patterns of data that accurately reflect the ground truth of the condition in question. The combination Etienne- Jules Marey foundational methodology of pulse wave analysis, captured across multiple regions of the face, across tens of thousands of individuals applied to modern biosensors, advanced AI and ML and we now truly have Dr McCoy’s tricorder capable of analyzing patients for their disease risks or vital signs measures.

The traditional methods of determining disease risk involve a trip to the healthcare provider, blood testing and ecg at the local lab followed by a review of results at a later date. Costly in both time and resources to determine if a patient is at risk of certain conditions. While this technology is not meant to replace medical devices or traditional methods of blood measurements, the implications for population health screening are incredible. The ability to capture disease risk analysis using the ubiquitous technology of smartphones now allows a platform to bring NCD (Noncommunicable disease) risk assessment and management to ½ of the global population without access to proper healthcare.^5.

Democratizing health literacy to help individuals better understand their health risks. There are 700 million people with hypertension globally, but nearly ½ of those individuals are unaware of their risk.^6.

We can now expect true population health by providing a platform for individuals to determine their risk of conditions like hypertension, diabetes, fatty liver and even mental stress measurement. Widespread screening would not be as feasible using the traditional methodology of medical office or laboratory visits let alone the willingness of patients to participate in screening programs. Bringing the assessment to the patient as point of care, or in defining equitable healthcare the point of need, will change the world and be a key ingredient to creating the global good of proper healthcare for all and achieving the SDG # 3 outlined by the UN.^7.

About Dr. Keith Thompson

Dr. Keith Thompson is a family physician and the Chief Medical Officer of NuraLogix, the company that created the world’s first contactless blood pressure measurement technology that measures over 30 health and wellness parameters using a conventional video camera that extracts facial blood flow information from the human face. 

References

1. Hong Luo, Deye Yang, Andrew Barszczyk, Naresh Vempala, Jing Wei, Si Jia Wu, Paul Pu Zheng, Genyue Fu, Kang Lee and Shong-Ping Feng. Smartphone-based blood pressure measurement using transdermal optical imaging technology. Circulation: Cadiovascular Imaging. 6 August 2019/https://doi.org/10.1161/CIRCIMAGING: 119.008857 
2. Deye Yang, Guomin Xiao, Jing Wei, Hong Luo. Preliminary assessment of video-based blood pressure measurement according to ANSI/AAMI/ISO81060-2:2013 guideline accuracy criteria: Anura smartphone app with transdermal optical imaging technology. Blood Pressure Monitoring 25(5):p 295-298, October 2020. / DOI:  10.1097/MBP.0000000000000467
3. World Development Report 2016: Digital dividends. Overview booklet. World Bank, Washington DC. License: Creative Commons attribution CC BY 3.0 IGO.
4. Zhou, W, Wang, Y, Gu, X, et al. Importance of general adiposity, visceral adiposity and vital signs in predicting blood biomarkers using machine learning. Int J Clin Pract. 2020; 75:e13664. https://doi.org/10.1111/ijcp.13664
5. Https://www.who.int/news/item/13-12-2017-world-bank-and-who-half-the-world-lacks-access-to-essential-health-services-100-million-still-pushed-into-extreme-poverty-because-of-health-expenses. December 13, 2017.
6. Https://www.who.int/news/item/25-08-2021-more-than-700-million-people-with-untreated-hypertension. August 25, 2021.
7. United Nations Department of economic and Social affairs- Sustainable development. Https://sdgs.un.org/goals/goal3.

biomarkers
medical
wellness
healthcare
imaging
health
device
devices
ai
machine learning
smartphone
software
fda
research