How this startup is using AI to predict chronic diseases

Finding doctors, discovering diagnostic centres, getting medicines and services isn't as difficult as it once was with several platforms, online delivery portals and consultation portals ensuring easy connectivity. However, in this world where healthcare connectivity is the priority of most startups, HealthNextGen uses artificial intelligence to predict diseases using advanced statistical methods in machine learning.

It predicts diseases utilising clinical parameters as well as other demographic variables. Their current disease model works for chronic disease complications such as cardiovascular diseases, diabetics’ retinopathy, sepsis complications and readmission risks after discharge from the hospital.

HealthNextGen also provides population health solutions that enable learnings for several diseases as well as the statistical prediction of the spread of certain diseases. Apart from predictive analytics, the platform ensures tracking of chronic patients ensuring medication adherence, communication with their doctors, and tracking their vitals to ensure that they stay healthy.

The beginnings

For 15 years now, Venkatesh Hariharan and Suresh Mallandira have known each other. In 2015 they began brainstorming on how they could transform the healthcare and treatment space in India. Santosh Hariharan, Venkatesh's brother, decided to bring in his experience as a healthcare researcher at Stanford Burham and the University of Toronto to HealthNextGen.

It began as a platform that provides practice management and electronic medical records (EMR). The trio had discovered that the biggest challenge in India is to ensure availability of EMRs to improve the quality of healthcare for the patients.

"Availability of EMRs will provide the data that our tools can use to predict the patient’s prognosis, and prioritises them for intensive treatment and tertiary care. We believe a legislation to mandate electronic health records in India will go a long way in improving patient treatment and reducing healthcare costs," adds 40-year-old Venkatesh.

The pivot and additions

The trio decided to build HealthNextGen into an intelligent platform that helps predict diseases as well. The team realised as the platform was clinical phenotypes oriented, it required a high degree of expertise from scientists and technologists. Their platform enables providers to provide personalised patient-centric care by tracking their progress throughout the treatment.

Venkatesh adds that their platform is scalable allowing them to incorporate several disease models, track population health at the institutional, regional and national levels.

Building a collaborative team

In addition to technology developers, the team also has a post graduate genomics researcher (PhD. -Genomics data scientist) on board, based out of Toronto.

"We are shortly adding a couple of medical doctors and business advisors to the board. Overall, we are a global team based out of Dallas, Toronto and Bengaluru getting together to make a positive impact in Healthcare," says Venkatesh.

Venkatesh says they have a few service-based competitors in US while there is no significant competition in India for their analytics platform. He adds that they hope to have early-mover advantage in healthcare predictive analytics and population health.

While there are competitors in India for their patient management platform, the team believes that their unique approach of predictions with an underlying platform with patient-level clinical information is well-differentiated in the market.

Revenue model

HealthNextGen has an SaaS-based model of revenue, charging per patient (for both EMR, adherence functionality and their predictive analytics to the providers).

For some providers, who already have advanced EMR, they are implementing predictive analytics on a per implementation basis based on number of patients, clinical history and data size.

Since implementing their new strategy six months ago, the team have rolled out their beta platform (HNG@Manage) to 15, 000 existing patients of their HNG@Clinic app.

What do the products do?

"We are in discussions with major hospitals in India (as well as the US) to implement our predictive analytics application on their population data. We are getting significant interest from these providers and hope to implement our technology over the next few months," says Venkatesh.

The two core products are:

1. HNG@Explore is their machine learning analytics application, which currently performs risk stratification using their algorithm along with the BI tool for graphical visualisation at the population health level as well as the patient level. "Our long term goal would be to further use patient’s medication data and be able to predict best combination of approved medications (and cost effectiveness) for any patient," says Venkatesh.
2. HNG@Manage - is the patient-level electronic health record. It also tracks patient medication adherence and gives reminders. Both products are on Microsoft Azure platform.

Future plans

This year, the team intends to integrate their product with major healthcare providers in India. They also have significant plans to expand and automate the collection of clinical and vital information from several sources including IoT (Internet of Things) device integration.

In addition, they plan to expand the research on additional disease models, and automate image-based (X-Ray, CT, MRI, cell based images) disease detection and predictions.

"We will be working with major providers in India and globally to facilitate data collection. We will also utilise our predictive learning techniques to improve healthcare and save healthcare costs through the shortest clinical pathway for diagnosis," says Venkatesh.

YourStory take

Data for 2000-2015 from the Department of Industry Policy and Promotion shows FDI investment in hospitals and diagnostic centres totalled $3.21 billion. According to IBEF, the healthcare market in India was worth close to $100 billion as of 2015, and is expected to touch $280 billion by 2020, clocking a CAGR of 22.9 per cent. Of the overall market, the healthcare delivery sub-segment accounts for 65 per cent.

Dr. Abdur Rub, co-founder of Xcode, a health genomics company, working on personalised, DNA-based wellness programmes, says “Some of the key trends seen in 2015 are deeper penetration of IT in various aspects of healthcare. There has also been greater awareness and adoption of preventive healthcare using technology-based devices by consumers,” adds Abdur.

HealthNextGen seems to be headed that way, but then it also has to compete with deeply funded startups. Practo has already ventured into the online ordering and e-commerce space; it also acquired Insta Health last year. Shashank ND, Co-founder Practo had told YourStory that they would be looking at data to enhance their platform.

Website