Download Request Form

Thank You

  • Close

Subscribe Now

Kindly fill in your email id

Thank you for your interest in YES BANK’s Savings Select Product. Our representative will get in touch with you shortly

All fields are mandatory.

Kindly fill in your contact details

Contact No should not start with 0

ESB Error

Thank you for subscribing!

Next News

Read More

Next Gen Healthcare: Artificial Intelligence for Patient Support

June, 2017

Ramakant Vempati, Co-Founder, Touchkin, on how Artificial Intelligence techniques are redesigning healthcare delivery

Kia is a 35 year old senior manager with diabetes, has mood swings, weight issues and is struggling to stay active, with a hectic work schedule and family to look after. She was diagnosed last year, and is often worried about the impact diabetes has on her quality of life and complications in the future.

Despite trying to follow the prescribed treatment and diet, Kia often finds herself at war with her moods, fatigue and increasing anxiety about her condition. She only sees her doctor once a month,  and frequently has questions about how to manage her mood swings or how can she manage an active life shuttling between work and family. But she has no one to speak to about this, and she stops caring about herself.  Her health slowly deteriorates, and she lands up in hospital with complications.

Does this sound familiar? Kia’s case is just one example of how the healthcare sector faces unprecedented demand for support from people who suffer from physical or psychiatric conditions.

NCDs: A New Paradigm of Care

Non-communicable or chronic diseases (NCDs) like diabetes and mental illness form the biggest burden of disease globally. The WHO estimates that one in four people globally will suffer from mental distress. India has only 5,000 trained psychologists for 1.2 billion people. People with diabetes, and those on the borderline, number close to 100 million.  The cumulative economic loss due to NCDs in low-middle-income countries is USD 7 trillion, nearly 500 times the cost of intervention.

The current healthcare system is ill-equipped to cope with NCDs. It is easier to manage acute or emergency situations as patients seek help when something happens, but it’s different for chronic diseases where patients need constant, personalised support. Short supply of trained staff, lack of knowledge and time, and the inability to be ‘always there’ means patients don’t get the support they need, or aren’t even aware that they need help.

This gap between demand and supply and improved access to technology points to another industrial revolution. Technology now reaches patients in remote areas, and tele-health is taking off.  Data Analytics and Artificial Intelligence (AI) techniques are redesigning care delivery ranging from diagnostics, case intake, to post treatment follow-up.

AI in Healthcare: Diagnostics & Genetics

What is Artificial Intelligence, or AI, in the context of healthcare? AI uses algorithms and software to approximate human cognition in the analysis of complex medical and patient data, and analyze the link between prevention or treatment techniques and patient outcomes. This is used to better manage patient treatment, and provide physicians with the information they need for a good decision.

AI programs have been developed for diagnosis, treatment protocol development, drug development, personalized medicine and patient monitoring & care. Medical institutions like The Mayo Clinic, Memorial Sloan Kettering Cancer Center, the National Health Service (UK), and firms like IBM and Google are using AI solutions in healthcare.

A popular use of AI is in diagnostics. Medical Sieve has built a “radiology cognitive assistant” to reason using broad clinical knowledge. Created using Watson, IBM’s AI platform, the assistant learns from images and existing diagnoses. It detects issues faster and with more reliability than humans, and is now qualified to assist clinical decisions for radiology and cardiology. Google DeepMind is analysing more than 1m eye scans in the NHS to train a neural network to identify early signs of degenerative eye conditions. Closer home, Bangalore-based SigTuple is using deep learning and AI powered diagnostics for automated analysis of lab samples, retinal scans, and x-rays.

Another application of AI is in the analysis of genomics. Deep Genomics combines machine learning and genome biology. Deep Genomics’ technology can do molecular diagnostics and screen risk assessments for complex disorders.  Their system models aspects of molecular biology and can be used for any disease or variant.

Next Gen Healthcare: AI for Patient Support

A relatively new adaption of AI for healthcare has been the creation of healthcare chatbots: as a virtual nurse, customer service manager, or a personal life coach.

Much like Whatsapp, users engage with chatbots in a chat window on a website or smartphone. Most applications have been in banking, travel, entertainment and hospitality in helping customers complete transactions or deal with issues. But the next frontier is healthcare.

At Touchkin, we’ve created one of the world’s first AI-enabled ‘Virtual Coach’ platforms for behavioural health.  The Coach combines insights from passive mobile sensing data with Natural Language Processing, empathy-based algorithms, and evidence-based therapy to create an engaging interaction. It also links to a real coach for more specific medical advice.

We think this is game-changing. Changes in attitudes and behaviour related to managing health and wellness are critical to achieve good outcomes, especially in managing chronic diseases like diabetes or conditions like depression. The Touchkin solution builds this support into common care situations, for people who may need motivation and changes in behaviour to support their health goals. It is designed to be ‘always-on’, inexpensive, and massively scalable. 

For example, Touchkin’s ‘sugar buddy' helps people with diabetes stick to an exercise and diet plan, overcome mental barriers (and excuses) preventing them from checking blood sugars regularly, or getting more active. With diabetes, this approach is being piloted globally in pain management, smoking cessation, and return-to-work. We are seeing 5-10X increases in patient engagement and staff efficiency, and 10-45% improvement in clinical outcomes.

How does the Virtual Coach support Kia? It helps her manage her diabetes in a structured way, track her mood, activity levels and progress. She now feels heard, connected with her own treatment plan, and confident about managing her condition. Every time Kia logs in, her Coach gives her some tips on how to live more effectively (and happily) with diabetes. She’s able to make better choices about food, overcome ‘no time for exercise’ excuses, and motivate herself when she’s feeling low. She also shares data with her doctor and diabetes-nurse, so they know how she is outside the clinic.

In a world facing an ever-increasing burden of chronic disease, AI-enabled Virtual Coaches can deliver proactive and personalised patient care at scale.  Here, we are glad to play a role.

The opinions expressed in the article are the author’s own.