Present - My brother, Vardhaan, and I are continuing to work on this project.
Low-Cost Cloud-based Contactless Vital Signs Monitor using PhotoPlethysmographic Imaging & Infrared Sensing Techniques
To design and develop a contact-free temperature and heart-rate detector & monitor that is not only low-cost but present many advantages like
Unobtrusive: They do not hinder movement or uncomfortable as in the case of wearable devices.
LifeSaver: remotely assess condition of patients/victims without risking the lives of Doctors & medical officers in contagious circumstances;
First response officers (paramedics, firemen, police) during hazardous circumstances
Hygienic: no need for consumables or sensor cleanings like traditional vital signs measuring devices or more recent wearable devices.
Notifications: Alerts and notifies guardians of any irregularities in vitals signs
Maintains Record: Logs information for future analysis
Vital signs (temperature, heart rate, respiration rate) are measurements of the body systems that monitor vital signs are among the most important tools healthcare professionals have at their disposal when looking after a patient. Even small changes to certain vitals can be early warning signs of a problem that can be detected long before any discomfort is felt by the patient. And in cases where patients are unconscious, sleeping, unable to communicate, or in a hazardous situation, vital sign monitoring is even more crucial.
Current systems in place do a good job of monitoring patients’ vitals and alerting medical staff of any abnormalities or causes for concern. But these systems all utilize sensors that require constant and secure physical contact with the patient. The few contactless monitors available are very expensive (thousands of dollars) and not commercially available for residential use or for use in developing countries.
My device is a temperature sensor and pi camera based vital signs detector using the principles of photoplethysmography, the process of deriving the measurement of blood flow through optical means. In order to obtain information about heart rate, the code superimposes a green light onto the forehead. This is because when the heart pumps the blood and contracts, it pushes an excess amount of blood in the arteries. When green light hits these expanded arteries, it will get absorbed at a faster rate than when the heart expands and a number of arteries decreases. This means that the rate of green light absorption is indirectly proportional to the heart rate. Information about the patient's temperature, heart rate, and respiration rate, along with an image of the patient at that given time. Thus, the guardian of the patient can easily view the patient and their vital sign.
The total cost of my prototype is $70. I tested it with 62 people, each person’s heart rate and temperature produced by the prototype was validated by an oximeter Pyle Health Bluetooth 4.0 Non-contact electronic thermometer. On each person, I conducted eight tests. I checked their heart rate and temperature 3 different times and to be sure I changed the thresholds for notifications to see if it would send out an alarm, so in total, I conducted 496 tests. My prototype had an extremely great accuracy for the temperature of 99%; however, the results of the heart rate were less than optimal. The rate for heart rate was that of 69% percent accuracy. After careful re-evaluation, I discovered many factors affect the end result. For example, the flickering of the lights may confuse the program into thinking that it is the heart rate. Other factors include frame rate, resolution, compression, image background, skin tone, motion, and lighting conditions. In the ambient and natural light, however, in the light that is constant, the accuracy increases by 10%.
My prototype can be used as a cost-effective commercial and residential solution for monitoring vital signs and alerting the guardian of any irregularities of the body. For future improvements, I want to conduct more tests with a constant background and noting skin-tone information, use Raspberry Pi Zero to lower the cost further, add a night vision camera to work into low light, look into possible extensions like sleep apnea and bruxism. Also, I would like to note that privacy is a key point in these kinds of devices, so I would like to add AES 128 and WPA2(WIFI) for the communication layer and the HTTPS protocol as a security measure from the tunnel between each sensor not and the web server.