The intent to build the BRIDGE Guideline was recently announced at HLTH. This guideline aims to reshape how healthcare systems approach AI integration at scale. It will focus on addressing long-standing challenges like system fragmentation and scalability, providing a comprehensive roadmap that helps healthcare organizations fully unlock AI’s potential.
We spoke with Josh Streit, AVP, Digital Transformation at Aidoc, and Brad Genereaux, Global Lead for Healthcare Alliances at NVIDIA, to dive deeper into the key challenges the BRIDGE Guideline will address, the strengths of this collaboration and the benefits for healthcare providers.
The BRIDGE Guideline will do more than set new standards – it will create an actionable framework that streamlines AI integration and enables real-world clinical impact, helping healthcare systems drive better outcomes for both patients and clinicians.
Want to be part of this journey? Receive exclusive updates on the BRIDGE Guideline and learn how your organization can get involved in shaping a framework set to transform AI adoption in healthcare – click here.
Why is now the right time for these guidelines, and what gaps are they addressing in the current healthcare AI landscape?
AIDOC RESPONSE:
Josh: The healthcare industry is facing a challenge right now trying to manage the 1,000 FDA-approved algorithms, and it will only get more difficult. One problem is with hundreds of different companies, each algorithm is developed independently, so there’s a lot of variation which makes it difficult for health systems to adopt these innovations in a practical way. The BRIDGE guideline aims to offer a solution – a neutral, streamlined approach to help healthcare systems concept and integrate these diverse technologies over time, in a way that’s manageable both in terms of time and cost.
NVIDIA RESPONSE:
Brad: Computer vision and generative AI have shown to be transformative in medical imaging – empowering radiologists and informaticists with insights to help in the triage, diagnostic and collaboration processes. However, the industry has hit the problem of enterprise scale. To build the sheer number of AI solutions necessary for things we can see in medical images – for example, https://gamuts.net puts that number at ~17,000 – multiplied by the number of hospitals and imaging centers in the world (likely exceeding more than 100K facilities), we need a new paradigm to democratize and help deliver on the transformative promise of these technologies.
What makes the BRIDGE collaboration unique?
AIDOC RESPONSE:
Josh: It is exciting to have two companies with different strengths from adjacent-industries working together to bring a unique point of view to the AI deployment challenge. NVIDIA has been providing the infrastructure and software needed to develop AI-driven tools in healthcare from the very beginning of image recognition. Aidoc, for the past eight years, has focused on successfully implementing these AI tools into clinical workflows across the globe. We’re building this guideline to help support the entire industry. We see this as a ‘rising tide’ moment. Our goal with the BRIDGE guideline is to drive both AI innovation and adoption, creating a practical, actionable roadmap that helps healthcare providers integrate cutting-edge AI solutions into their workflows and ultimately elevates the entire healthcare ecosystem.
NVIDIA RESPONSE:
Brad: Other industry efforts are addressing other critical parts of the problem. There are development frameworks – key amongst them MONAI – that are helping solve the world’s need for a ubiquitous mechanism for developing AI. There are standards bodies crafting the API specifications and profiles to connect these standards together. Regulatory bodies have put together frameworks to assess the safety and appropriateness of these AI solutions and how they are used. What this collaboration does is puts together a comprehensive set of guidelines to help ensure best practice approaches in packaging and deploying AI solutions in hospitals, mitigating scalability issues.
What do you see as the primary obstacles hindering the transition from idea/initial product into practical adoption at the clinical level and how will the guideline work to address this?
AIDOC RESPONSE:
Josh: Solving this is the central aim of BRIDGE. The simple arithmetic of the labor challenges in healthcare means that today’s clinician can benefit immensely from productivity enhancements. This places significant emphasis on the need to infuse their clinical workflows with the power of AI, helping them adequately reach, react, and respond to the volume of patients in need of their care. BRIDGE is a universal guide intended to cover both the creation and implementation of those tools in a standard manner. Without a standard, unifying guide on the effort needed to reach production adoption within a health system today, we have observed many hundreds of solutions get produced with only dozens reaching utilization and scale. This is not an adequate enough enhancement to enable the typical clinician to reach the additional patients in need of their care and attention. We can use the adoption of AI-driven tools to go much faster. To do this, we must be efficient. To be efficient, we must be able to implement new tools in a predictable and cost-effective manner. This is the intent of BRIDGE for everything from model creation to clinical product adoption and its drift mitigation.
NVIDIA RESPONSE:
Brad: The primary obstacle is the sheer amount of variability we find in the world today. There are many systems, people and workflows, and so much customization involved in building a cutting-edge algorithm for healthcare systems. It’s extremely difficult to scale if every AI application is developed in isolation, and delivered on standalone infrastructure. This guideline will give to those on the frontlines and those building solutions, a common recipe and common set of expectations, to streamline the work that they do and to reduce the number of exponentials. This will help build toward more impactful, resilient AI solutions at scale and with resilience.
What does the future of AI in healthcare look like?
AIDOC RESPONSE:
Josh: The future is a multi-modal, fast-paced era of pluralistic participation of industry and clinical experts, vendors, hyperscalers, scientists and innovators from around the world, harnessing their collective experience and ingenuity to assist clinicians and enhance healthcare systems. Legacy technology is struggling to deliver to clinicians the productivity enhancement they need in order to keep up with the volume of patients under and in need of their care. Working with NVIDIA, Aidoc sees its role as the manner by which customers, clients and partners can enable their creations to reach production use. We believe it is a very exciting time in which we can help create and foster a community of users whose ingenuity may help each and every one of us someday as these tools proliferate across each enterprise.
NVIDIA RESPONSE:
Brad: Digital agents are set to assist in all parts of healthcare helping radiologists, patients and informaticists alike. These agents are digital twins that reflect the workflows and insights needed to better the health, the experience and the treatment overall journey of the patient. To make this vision a reality, NVIDIA is building and accelerating the systems and frameworks to craft these digital twins, enabling the ingestion of signals and delivering insights to those that need them.
