Blockchain technology is gradually becoming part of how we think about healthcare data. Even government entities like the ONC and FDA – typically not early adopters – are throwing their hat into the blockchain ring.
In fact, according to recent research by Deloitte, healthcare and life sciences companies are planning the most aggressive blockchain deployments of any industry. Thirty-five percent of Deloitte’s respondents told the consulting firm that they expected to put blockchain into production this year.
Many companies are tackling the practical uses of blockchain tech in healthcare. But to me, few are more interesting than Google’s DeepMind, a hot new AI firm based in the UK acquired by Google a few years ago.
DeepMind has already signed an agreement with a branch of Britain’s National Health Trust, under which it will access patient data in the development healthcare app named Streams. Now, it’s launching a new project in partnership with the NHS, in which it will use a new technology based on bitcoin to let hospitals, the NHS and over time, patients track what happens to personal health data.
The new technology, known as “Verifiable Data Audit,” will create a specialized digital ledger which automatically records every time someone touches patient data, according to British newspaper The Guardian.
In a blog entry, DeepMind co-founder Mustafa Suleyman notes that the system will track not only that the data was used, but also why. In addition, the ledger supporting the audit will be set to append-only, so once the system records an activity, that record can’t be erased.
The technology differs from existing blockchain models in some important ways, however. For one thing, unlike in other blockchain models, Verifiable Data Audit won’t rely on decentralized ledger verification of a broad set of participants. The developers have assumed that trusted institutions like hospitals can be relied on to verify ledger records.
Another way in which the new technology is different is that it doesn’t use a chain infrastructure. Instead, it’s using a mathematical function known as a Merkle tree. Every time the system adds an entry to the ledger, it generates a cryptographic hash summarizing not only that latest ledger entry, but also the previous ledger values.
DeepMind is also providing a dedicated online interface which participating hospitals can use to review the audit trail compiled by the system, in real-time. In the future, the company hopes to make automated queries which would “sound the alarm” if data appeared to be compromised.
Though DeepMind does expect to give patients direct oversight over how, where and why their data has been used, they don’t expect that to happen for some time, as it’s not yet clear how to secure such access. In the mean time, participating hospitals are getting a taste of the future, one in which patients will ultimate control access to their health data assets.