This story is over 5 years old.

A 'Cognitive Healthcare System' is Changing Medical Practices in Remote Areas

A "cognitive healthcare system" links rural patents with doctors, making medical care accessible for even the most remote patients.

In rural communities throughout the developing world, a doctor can be nearly impossible to find without a long, prohibitively expensive journey to a city. Because of this, many people around the globe have a difficult time accessing crucial healthcare services. For example, in India's villages, many fall victim to preventable diseases, according to the Ministry of Health, while two-thirds lack access to critical medicines.


This was the problem a team of engineering students in Pune, India, set out to solve. Their project, IoT Vaidya, is a "cognitive healthcare system" that links rural patients with doctors, wherever they might practice, using sensors and cloud technology. Recently, IoT Vaidya won the Intel® Ultimate Coder Challenge for IoT, in which five teams were given an Intel® IoT Developer Kit and eight weeks to design an innovative commercial IoT solution.

Motherboard spoke with team co-founder Mahavir Dwivedi about their winning project and the healthcare challenges facing impoverished rural communities in India.

Motherboard: Can you tell me a little about your background, and how you met up with your team?

Dwivedi: I graduated in electronics engineering from Pune University in India last July. Since my sophomore year, I've been working on projects which could hopefully turn into real commercial products. Arun Kumar was a classmate of mine; we have done several projects together. He has special interests in robotics and machine learning, and also just graduated this year.

Milind Deore is our mentor. He works with CISCO as a network engineer. He has more than a decade of experience in hardware and software. Beyond the technical approach, he's guided us towards projects that can be launched into the market.

A series of sensors transmit patient information using cloud infrastructure for remote diagnosis by medical professionals.

How did you come up with the idea for IoT Vaidya?

Before participating in the Intel® Ultimate Coder Challenge, we had been to various hackathons and entered a few competitions. Though our performance was fair, we never could take those projects to the next level. We decided we wanted to work on something that could be useful to society, which is how we came to focus on healthcare. It's a real issue here in India.


After doing a substantial amount of research, we learned that though the market is gearing up for a biometric wearables revolution, countries like India still suffer from a critical shortage of doctors. The situation is particularly bad in rural areas. 70% of the Indian population has access to less than 20% of physicians, 10% of the specialists, and 5% of the diagnostic equipment. The situation is not much different in other developing nations in Asia, Africa and Latin America.

Thus, we had a problem to solve. We decided to build a standalone cognitive health monitoring system, IoT Vaidya, which provides real-time diagnoses to patients, and connects them with doctors and specialists using cloud infrastructure and telemedicine-enabling technologies.

For a name, we chose to use the Hindi term for doctor: "Vaidya."

How does it work?

The project uses biomedical sensors to acquire a patient's vitals, like blood pressure and temperature, which are then sent to the cloud where they can be accessed by a physician. This way, a doctor working far away, in any corner of the world, would be able to diagnose a patient's problem. We have also written machine learning algorithms in the cloud for a very basic analysis of the data, so that instant feedback can be given back to the patient regarding preventive measures to be taken.

Machine learning algorithms aid in diagnosis.

How long did it take to develop?

This was not the first competition we've entered using this healthcare-focused solution. We won a previous competition, the Eclipse Open IoT Challenge, with this same project. However, in that case, we were not given Intel boards, so we used open source technologies, like Arduinos. You can use those boards to build prototypes pretty easily, but in order to bring something to market you need certain things only the big players can provide, since they're more acceptable and tested.


So we brought that prototype to the Intel® Ultimate Coder Challenge for IoT, which is when we got Intel-backed software and hardware, like the Intel® Edison board and Intel® IoT Developer Kit. It took us around six weeks to import our previous work on to Intel platforms, and come up with the further developments.

What programming languages did you use in writing the software?

We didn't limit ourselves to any one particular language or platform. As this was only an eight-week competition, it was not possible for us to write each and every bit from scratch. So we developed our solution using various open-source libraries and frameworks. In general, we used Python for writing scripts on Edison, C++ for writing Arduino programs, and JavaScript on the cloud.

What was the most challenging part of the project?

The most important part of our project was the sensors. The Intel® IoT Developer Kit included some amazing sensors from Grove, but it was never going to be enough for us to build a healthcare monitoring solution. So we included our very own. We consulted doctors to get acquainted with the type of data they required for primary diagnosis of a patient, especially for common things like heart disease. All the sensors we considered using had to be non-invasive, so that infections could be avoided. Ultimately, we settled on four: galvanic skin response, temperature, blood pressure, and electrocardiogram.

What's the next step in developing this project?

During the last six months, we participated in various competitions and developed the proof-of-concept of our product. Now, we are collaborating with other healthcare startups, and have started working on taking this concept to the clinical trials stage. We are also working on our algorithms to deliver more accurate diagnoses to patients. It's my hope that this will prove useful not only in India, but in rural areas lacking basic health care throughout the world.