Tech-led fitness startups are using Artificial Intelligence (AI) systems to alert users about potential lifestyle conditions early enough for accurate diagnosis, in the process perfecting a business model that focuses on prevention for the health conscious. Typically, a fitness app on your phone looks at an Instagram picture of what you ate to gauge the calories consumed and suggest the best workout next day. Initially, firms such as GOQii, HealthifyMe and Cure.Fit hired dieticians and fitness experts to manually scan each image and come to accurate conclusions.
AI: Designing exclusive packages
These days, however, a bulk of this grunge work is automated through AI systems. These firms are trying to mimic human intelligence, specifically fitness coaches, dieticians and even doctors, through their products. The benefits are plenty. AI systems are infinitely scalable, so these companies can serve millions with relatively little manpower. They are also far better than humans in identifying patterns and learning from millions of data points. Their accuracy will only improve as sensor technology gets better.
Watches that keep your health record
So, from mere fitness suggestions, these companies through their apps have moved on to highlighting lifestyle conditions such as diabetes, hypertension and obesity. "Between your mobile phone and wearable, you are able to get a lot of health parameters live. The job of a service like GOQii is to use this data and, along with AI, make predictions to warn people of a catastrophic health situation down the line," says Vishal Gondal, cofounder of GOQii. Today, GOQii's wearable can track blood pressure and heart rate. An ECG functionality is also expected soon. The company says it tracks around 300 parameters in any particular user daily, translating to thousands of instances in just one month. (Pic: GOQii)
Stepping up the fitness quotient
This abundance of data is what sets these new-age technology companies apart, says Shamik Sharma, technologist at Cure.fit, which runs over 100 fitness centres across India. While traditional diagnostic labs get granular data about a user's health sporadically, AI becomes key in preventive healthcare as it can build mechanisms which capture a lot more immediate data, he says. "It's not only that someone wants to lose 5 kg so they come to us; people are now starting to use us with the goal of better controlling diabetes and other (lifestyle) diseases." India's sheer numbers give companies working on such solutions enough data points to accurately train AI systems. (Pic: Cure.fit)
When AI plans your diet
"Nowhere would I have been able to hire 500 nutritionists and trainers like we have today. Granted, they're not building the AI, but they've indirectly trained the system because the data used is from the chats they've had with our customers," said Tushar Vashisht, cofounder and CEO of HealthifyMe. HealthifyMe's AI-based fitness assistant Ria can automatically suggest diet plans to a user, which may seem quite straightforward, but the algorithms take into account thousands of parameters. The firm gets more data on users suffering from specific diseases, which are then leveraged to help others suffering from a similar fate.
Let's discuss insurance
For now, these digital fitness services operate on a monthly or yearly subscription model, but they could start eyeing additional revenue streams from data sharing in future. GOQii has partnered with Max Bupa Health to enable its users to avail lower insurance premiums based on their fitness data. Data sharing with hospitals too is possible in the near future, with doctors getting a better picture of a patient's health based on diet and fitness data.
Big companies, bigger plans
While several fitness companies across the world are trying to use AI for preventive healthcare, successes haven't been all that common. It's largely been giants such as Apple, which has had a big headstart in training its AI systems, that is seen leading the way. Apple and Google have fitness apps baked into the iOS and Android operating systems, but have so far stayed away from building prescriptive services that nudge users to stay healthy. Instead, these tech giants as well as Microsoft are building solutions to read scans, images and slides to diagnose a disease more accurately than doctors, for conditions such as diabetic retinopathy, cancer and cardiac issues.
AI and its limitations
Though AI is emerging as key pivot in preventive healthcare, its utility ends at nudging users to eat healthy and work out more often, or suggesting they visit a doctor. No AI can be prescriptive, as of now. Cure.Fit, for example, gets doctors to see a user's fitness and diet data in its primary care centres, but a case of AI replacing the doctor for the most basic diagnosis is far in the future. "AI being used for self-monitoring by looking at your lifestyle and purchases is probably the first place the technology will make an impact in healthcare at scale," says Thulasiraj Ravilla, director of operations at Aravind Eye Hospitals, which worked with Google to train an AI system that can screen patients for diabetic retinopathy.