So you’ve managed to convince everyone that you’ve kicked the habit while continuing to light up furtively behind their backs? 

Well, artificial intelligence can catch you red-handed, or rather “smoker-faced”!

Computer vision techniques are used to identify signs associated with lifestyle factors like smoking, which causes darker lips and vertical wrinkles rather than the horizontal wrinkles seen in a non-smoker.

These indicators collectively form what is known as ‘smoker’s face’, says Prasanth Madavana, co-founder of health-tech startup Fedo.ai.

Data-led diagnosis

The company uses AI and machine learning (ML) technology to map 14 physiological parameters, pinpoint potential disease risks, and estimate the likely medical expenses for the next five years — all through a 14-second facial scan on a mobile phone app.

Nearly 70 per cent of all health risks are lifestyle-related, mainly through poor habits, Madavana says, citing a WHO study. “Therefore, we will first focus on non-communicable or lifestyle diseases, given that early detection becomes crucial to insure one’s health,” he adds. 

The parameters analysed include body mass index (BMI), heart rate, blood sugar level, heart rate variability (HRV), cholesterol level, respiration rate, stress level, electrocardiogram (ECG), blood pressure, oxygen saturation level, and haemoglobin level. 

Additionally, Fedo.ai takes into account factors like smoking, inhaler use, and medications to compute a score ranging from 0 to 1,000, with 1,000 indicating optimal health. 

“An average person in good health would typically score 600 or above,” Madavana says.

Healthy roadmap

Additionally the app deploys algorithms for disease risk prediction for conditions such as type 2 diabetes, hypertension, obesity, cardiovascular disease, coronary heart disease, chronic kidney disease, and respiratory disorders. 

These disease risk predictions, aimed at facilitating proactive health management, are at a warning or directional level rather than clinical, with 80 per cent accuracy, Madavana says.

Fedo.ai uses contactless photoplethysmography (rPPG) and ML technology to measure blood flow and heart rate remotely. This involves measuring blood flow underneath the skin using ambient light. 

Funding

Integrating health services and technology, health-tech is a steadily growing market segment in India that is expected to touch $25 billion by 2025, according to a report by investment banking consultancy LoEstro Advisors.

Fedo.ai has raised funds from Unicorn India Ventures, SEA Fund, and former Max Bupa managing director and CEO Ashish Mehrotra. 

The startup’s clients include Aditya Birla Sun Life Insurance, Canara HSBC Life Insurance, Aviva Life Insurance, Aegon Life Insurance, Pramerica Life Insurance, and COCO by DHFL General Insurance. 

It is also working alongside hospitals and government organisations to collect data to improve AI accuracy.