Unlocking the Full Potential of GenAI

Insider News for August 2023

If it seems we’re hearing more about the potential of artificial intelligence (AI) to transform healthcare, it’s because each new application makes other applications possible. That’s how innovation works, exponentially elevating potential with each new discovery or invention. From better risk prediction for disease stratification to faster clinical trials to improved imaging diagnoses and better outcomes, AI holds great promise for advancing healthcare in ways never imagined just a few years ago.

To fully realize the potential of AI, machine learning (ML), Large Language Models (LLMs), and other innovations, we’ll need an effective deployment pathway. And that deployment pathway requires a new kind of interoperability and a new type of security. Imagine the effort it would take to deploy all these great new technologies using hundreds of data interfaces—not to mention trying to protect the vast amount of personal data being shared via these technologies.

Generative AI (GenAI) is a type of artificial intelligence that can generate text, images, and other modalities in response to prompts. GenAI solutions are built using LLMs trained to learn patterns and structures in large data sets and then predict and generate responses. But acquiring a significant volume of domain-specific data to train these models can be challenging.

Today, centralized data aggregation is the primary model for collecting training data for GenAI solutions, which comes with significant privacy and security risks. This approach is especially challenging when proprietary or sensitive data is used, and it requires considerable safeguard measures during model training and solution deployment. Unfortunately, this can limit the collective value of GenAI solutions. This approach also provides little to no transparency of how the data is used to train the models or how the generated data is eventually shared with users.

All these challenges can be addressed through Avaneer Health’s decentralized network model. The Avaneer Network can be a powerful distributed GenAI enablement platform to accelerate adoption:

Solutions have permissioned access to large, distributed, real-time data sets from network participants.
The platform is designed to enable patient consent on how data is used throughout the network.
The distributed solution model allows strong AI governance to determine what data can be used for what purpose (ex., Which data elements can be used for training).
Network traceability and auditing provides a detailed understanding of the data and data transformations used in the modeling process.
Solutions are deployed where the data resides (in the participant’s SparkZone^TM—a private, secure, cloud-hosted “home base” on the Avaneer Network where payers, providers, and other network participants can collaborate and transact directly with each other), allowing solutions to be trained to generate responses based on custom data sets.
Cloud GPU-based SparkZones can be configured to power model generation at the edge.

Initial industry use cases are focused on helping physicians with their clinical documentation, such as transforming conversations into prescriptions, follow-up appointment letters, and consultation summaries. However, the potential use cases are practically limitless and can advance granular patient consent.

Unlocking the full potential of GenAI

We can all agree that AI has the ability to transform patient experiences, improve outcomes, and significantly reduce administrative burdens. But truly realizing these benefits requires a new kind of network through which these innovative technologies can be leveraged. That network is here today.

We have exciting news coming on how we’re advancing innovation in AI. Stay tuned!

Gerald Bortis, PhD
Chief Technology Officer
Avaneer Health

In This Issue

Streamlining Workflows and Improving Access
WHO and HL7 Collaborate on Global Interoperability Adoption
The Evolution of Interoperability in Healthcare: Where We’ve Been, Where We Are Today and Where We’re Going
Patients Need More in Digital Patient Experience, Patient Access
AI and Large Language Models: The Future Of Healthcare Data Interoperability
Why It’s Critical to Emphasize Data Protection in a Digital World

Industry News

The average provider completes 41 prior authorizations every week for each provider in the organization, which equates to nearly two full business days. Four out of ten providers say their organization has had to hire staff specifically dedicated to managing the highly manual, time-consuming process. – American Medical Association

Streamlining Workflows and Improving Access

Outcomes Rocket Podcast
In this podcast, Stuart Hanson, Avaneer Health CEO, discusses the challenges of interoperability and why we’ve struggled to achieve genuine data fluidity after so many years of working on the problem. One of the challenges, according to Hanson, is that organizations have invested so much in legacy systems that it can be hard to admit that there might be a more effective way of doing things. And it’s difficult to be innovative when the vast majority of your IT budget goes toward keeping the lights on, maintaining connections, and supporting infrastructure. According to Hanson, “We need to take a realistic look at things from the consumer’s perspective and really recognize that there has to be a better way.”

Listen Now

WHO and HL7 Collaborate on Global Interoperability Adoption

Healthcare IT News
The U.S. isn’t the only country focused on advancing interoperability. The World Health Organization (WHO) recently signed an agreement with Health Level Seven (HL7) International to “jointly develop guidelines and advance the adoption of open interoperability standards” that are vital for “equitable and evidence-based digital health.” This informative article provides a high-level overview of the initiative, the countries involved, and the role of FHIR in global healthcare.

The Evolution of Interoperability in Healthcare: Where We’ve Been, Where We Are Today and Where We’re Going

HIT Consultant
Why is it that we can send money to nearly anyone around the world regardless of where they bank, but we still rely on CDs, printed copies, and faxes to share test results between two providers? The financial services industry has had data fluidity figured out for years while healthcare still struggles. Of course, healthcare is more complex. But even beyond that, we find ourselves tied to behemoth legacy systems that make data sharing and collaboration extremely difficult—even with FHIR.

In this article, Avaneer Health CEO Stuart Hanson explores these challenges and why innovation is the key to getting us “across the finish line” of interoperability.

Patients Need More in Digital Patient Experience, Patient Access

Patient Engagement HIT
Only 17% of patients recently surveyed say the digital patient experience is getting better, and many say it’s actually gotten worse since the pandemic. Booking appointments, making payments, long wait times to see a specialist, and patient navigation could all be significantly improved through digital tools—the same tools that can help reduce administrative burdens for providers.

This insightful article delves into the challenges that have caused the digital setback and what needs to happen to get the ball moving forward again.

AI and Large Language Models: The Future Of Healthcare Data Interoperability

Forbes
Data standardization is essential to achieving interoperability but is still limited in many ways. This forward-thinking, eye-opening article explores the limitations of current data standardization tools and introduces the potential of AI and LLM to fill those gaps. For example, LLMs can “organize and contextualize raw data,” something current standardization tools cannot do. “LLMs have the ability to read doctors’ notes in a matter of moments and organize the data to support easier analysis.” The capability extends to raw data of any kind, such as labs, clinical notes, and more. The potential to transform healthcare as we know it is immense—elevating interoperability to an entirely new level.

Why It’s Critical to Emphasize Data Protection in a Digital World

HealthData ManagementM
illions of patients in the U.S. have had their personal information exposed since the beginning of 2023 due to multiple data breaches. As more hospitals and health systems employ tools to enable electronic transmission of data between entities, they must be increasingly diligent in their efforts to keep that information safe. The vast number of APIs a typical hospital uses makes this challenging since this is where many hackers focus their efforts. This timely, important article provides proven tactics to do just that.