How an Indian whiz kid's AI solution aims to beat the organ donation racket

Pratik Mohapatra, a student at Bangalore’s RV College of Engineering built OrganSecure – an AI-based app that connects organ donors to recipients. Furthermore, with blockchain coming into play, the app will ensure a secure, tamper-proof database.

pratik mohapatra organsecure
Pratik Mohapatra

Four lakh people die every year waiting for an organ transplant. The demand-supply gap in organ transplantation in India is humongous. Take for instance liver transplants – for two lakh livers required, a mere 708 have been donated.

The waiting period for an organ can run into several years, and this is precisely why a thriving black market for organs exists in the country. The ‘waiting list’ for organs is primarily susceptible to manipulation simply because the process is not transparent and therefore, not 100 percent tamper-proof.

Influenced by Breathe, a web series that brings the organ donation racket to light, 21-year old Pratik Mohapatra researched online and found that in addition to a clear shortage of organs in India, there are numerous loopholes and problems in the process.

That’s when Mohapatra, a student at RV College of Engineering and a Microsoft student partner thought: How about using a technology solution to solve the problem? And it’s this thought that led to the creation of OrganSecure – an AI-based app that ensures complete transparency and makes it easier for the recipient and the donor in the organ donation process.

Pratik Mohapatra, Architect of OrganSecure Pratik Mohapatra

"Blockchain is a technology in which transactions or entries cannot be deleted. This ensures any tampering attempted on the ledger will be recorded. So users alone have complete ownership of the data."

-Pratik Mohapatra

Building OrganSecure

OrganSecure bagged the third spot in ‘Microsoft’s AI for Good’ challenge – a global competition that aims to recognize technology solutions for solving humanitarian issues and creating a sustainable world.

“The technology is currently in the alpha stage, so the full-fledged application is yet to be built. I used Microsoft Azure Machine Learning Studio – this gave access to the complete set of Azure products and services,” explains Mohapatra.

He used the regression model to help predict the probability of an organ matching the recipient. Simply put, linear regression establishes a linear relationship between one or more independent variables and a numeric outcome, or a dependent variable.

To create OrganSecure, Mohapatra used the Microsoft Azure virtual machine for Content Delivery Network, Azure IoT hubs, Azure Service Centre and Event Hubs. By using servers closest to each other, Azure Content Delivery Network can send files in a faster and more reliable manner.

So how was OrganSecure trained?

“In addition to using data from government websites, we also procured the data from hospitals. An aggregation of all this data was used to train the initial machine learning algorithm,” says Mohapatra.

Although the data needed to feed the app was quite limited, he explains that the model was trained on a mixture of data. Now the good part about randomized data is that one can add any test case and more often than not, it will be covered in data procured from hospitals – this, therefore, results in fewer ‘outliers’ in the data pool.

A key challenge Mohapatra is faced with is that India doesn't have a centralized database, it only has a gist of the information – not something that could be used to train the data model. 

However, given the fact that India has but a handful of organ donation apps, and none leveraging AI, Mohapatra is optimistic about more number of people signing up with OrganSecure, thereby making its data pool richer.

With the possibility of numerous recipients and donors signing up with OrganSecure, there could be a need to scale up rapidly. To futureproof the app for scalability, Mohapatra has opted for a solution that helps build and scale solutions using micro batching. 

Leveraging blockchain to take on the organ donation black market

As is the case with most commodities, the massive gap in demand and supply creates a thriving organ donation black market. 

To eliminate any possibility of the system being manipulated or tampered with, Mohapatra has set his sights on blockchain and smart contracts. “We want to move to a system that is blockchain-enabled. So all the information is stored on a ledger and is based on smart contracts,” he explains.

He adds that blockchain is a technology in which transactions or entries cannot be deleted. This ensures any tampering attempted on the ledger will be recorded. The users alone have complete ownership of the data.

Information on OrganSecure can only be accessed by doctors and patients. In fact, the doctor can only access the information if a patient gives the approval – the distributed ledger system puts the power entirely in the hands of recipients.

The next leg for OrganSecure

Being a firm believer in the ‘quality over quantity’ principle, Mohapatra says that OrganSecure will be launched only after it passes stringent quality checks.

"A lot of people have asked why I haven’t yet released the app in the market. However, for something as critical as organ donation at stake, the app would need to go through hundreds of thousands of tests. Only then would it be released into the consumer space," he shares.

Bangalore is the first city OrganSecure has set sights on; however, since the app has been future-proofed, Mohapatra believes scaling it to different regions shouldn’t pose a problem.

Copyright © 2019 IDG Communications, Inc.

9 steps to lock down corporate browsers
  
Shop Tech Products at Amazon