There's been a significant amount of momentum building for the use of blockchain technology in Life Sciences and Healthcare. For example:
- A Case Study for Blockchain in Healthcare. "MedRec" prototype for electronic health care records and medical research records. [LINK]
- Blockchain and Health IT: Algorithms, Privacy and Data. [LINK]
- Healthcare: Can the Blockchain Authorize it and Secure it? [LINK]
- Blockchain Applications for Healthcare [LINK]
- US Government Announces Blockchain Healthcare Contest [LINK]
With all the buzz it was a great time to attend an all-day Life Sciences and Healthcare Blockchain Workshop at MIT Media Lab. The workshop was jointly hosted by the MIT Media Lab and Deloitte. Deloitte has done quite a bit of work to develop their own blockchain point of view, including a blockchain lab.
In preparation for the workshop there was a significant amount of pre-reading and homework. Each company that attended the workshop needed to bring blockchain questions and opportunities from their own companies. In my case, having worked for Dell Technologies for a total of 4 business days, I was pleased to find that DELL accepts bitcoin payment through a partnership with Coinbase (I plan to dive deeper into the details behind our implementation in a future post).
The morning session was primarily focused on MIT Media Lab activity in this space. The afternoon session was dedicated to a Deloitte-led brainstorming session on the use of blockchain in healthcare from the view of provider, payer, insurer, patient, etc.
As an attendee I was quite interested in "working code" and spent most of my focus understanding the details behind two of the Media Lab's projects: MedRec and Opal/Enigma. In this post I'd like to cover MedRec.
- Unify patient access to their electronic medical records (EMRs) WITHOUT requiring the centralized gathering of medical records into one mega-repository.
- Enable patient-initiated data sharing of their EMR
- Enable big data research on EMR
The system architecture for MedRec (described here) is displayed below:
This architecture (which is deployed in prototype form with a local provider) does not store EMRs to the blockchain but instead stores state changes to a patient's EMR. The EMR data itself is actually stored into the provider's database and a pointer to it (along with permissions) is stored in the blockchain. Ethereum is a big part of the architecture using its smart contract feature as a way of enforcing and implementing policies.
The prototype uses 5 virtual machines to pull this off:
- A miner VM to generate the blockchain
- The provider's blockchain client VM, which constantly crawls the blockchain
- The patient's blockchain client VM, which also crawls the blockchain
- A VM that runs a webapp for the provider
- A VM that runs a webapp for the patient
During the workshop Ariel shared some ideas and future directions about how this architecture could be leveraged for Big Data and Analytics. Successfully implementing any of these would accomplish the third goal of MedRec (enabling analytic research on EMR):
- Leverage queries across the blockchain so that apps can show trends of your own medical records (e.g. year-to-year trends of symptoms felt during the same time each year)
- Predictive analytics based on MedRec population data
- Enabling researchers focused on precision medicine, defining their cohorts and specifying demographic data for longitudinal studies
- An analytics layer for trend discovery, narcotics abuse, conflicting prescriptions, epidemiological trends, etc
The project details fueled a large amount of questions from the medical community in attendance, and by day's end the general enthusiasm for the use of blockchain for patient benefit remained intact (if not increased)!
Dell Technologies Fellow