How private healthcare data can predict diagnoses, unlock innovation

The exchange of private data between organizations can foster rapid innovation in health care, especially through the ability to develop AI using previously unavailable data. However, medical facilities now face many obstacles when it comes to unlocking data.

By collaborating and sharing data without compromising the confidentiality, speed, and integrity of data, health care providers can predict future diagnoses and reduce the complexity of internal and external data sharing, which often includes confidential personal information.

“The most significant barrier to data sharing in healthcare is the high cost and high level of effort required to maintain compliance,” said Ridhiman Das, co-founder and CEO of Tripleblind.

Das, who will share his views on the benefits of private medical data next month at HIMSS22, said that today more than 130 jurisdictions have data privacy rules – these include HIPAA in the US, GDPR in the EU and PDPA in Singapore.

“Even in the U.S., there are many rules, such as the California CCPA,” he added. “Another barrier is that some health systems see themselves competing with other systems, making them less likely to share data.”

TripleBlind blind learning technology allows you to use private datasets to learn models without moving or re-identifying them. The data remains available to scientists on data, while complying with all privacy rules.

From his point of view, there are two main reasons why gaining access to private data accelerates healthcare innovation.

First, health-oriented algorithms, such as diagnosis and decision support, need to be trained – and the more diverse the data with which this training is conducted, the more accurate the algorithm will be.

“Today, health systems train their algorithms on local patient data,” Das explained. “Data from one system may lean toward older, whiter, sicker patients. If this system can access the data of colored people, young and usually healthier patients, the algorithms they will develop will be more accurate. ”

In addition, Das said, it used to be difficult to share certain types of data, such as genomic information, while maintaining patient privacy.

“Effective data sharing of this type opens up new opportunities for new diagnoses and treatments,” he said.

Ridhiman Das and his colleague from Triplelind, Mayo Clinic doctor Dr Suraj Kapa, ​​will explain more in their HIMSS22 session “Unlocking Private Medical Data”. It is scheduled for Tuesday, March 15, from 3pm to 4pm in Room W311E of the Orange County Convention Center.

Nathan Eddie is a freelancer in healthcare and technology based in Berlin.
Write to the writer: nathaneddy@gmail.com
Twitter: @ dropdeaded209

Leave a Comment