Accelerating GenAI Transformation: Highlights from the BCG Session on Generative AI's Role in Shaping the Future of Health Care

As the revolutionary potential of GenAI becomes more and more apparent, how can health care companies accelerate their GenAI journey? Parminder Bhatia , Chief AI Officer at GE Healthcare, John Doyle , Global Chief Technology Officer for Healthcare and Life Sciences at Microsoft, and Jean-Philippe Vert Chief Research and Development Officer at Owkin, joined me on stage during the J.P. Morgan Healthcare conference to share their thoughts on this critical question.

Selecting and Validating Use Cases

2023 was the year of experimentation, as health care companies focused on understanding what Gen AI is and what it can be used for. We saw, for example, that GenAI can accelerate development tremendously, getting products into the market in one year instead of five or ten.

The question companies face now is how to determine the highest-impact use cases—which ones will provide the greatest value for the organization in both the near and long term.

So far, automation has been the top health care use case. For example, software development has been accelerated by about 50 to 55%, allowing people to spend more time developing better code. In 2024, use cases will likely focus on operational efficiency—improving better bedside prediction, streamlining the scheduling, and so on. The next step will involve moving beyond large language models to multimodal models—AI models that use imaging, text, speech, and so on—and exploring those opportunities to make meaningful impact in health care.

Whichever use cases are selected, companies need to make sure that models are validated robustly to ensure data privacy and minimize bias. One panelist noted that 10% to 20% of his company’s AI development is actual development, and 80% is multi-site validation, making sure it's validated at least 20 sites before it's deployed. ?To ensure that every use case is validated in the same way, companies should consider creating a framework for use case ideation.

Implementing Use Cases at Scale

After companies have selected and validated the key use cases, how do they scale them quickly? Three enablers are crucial. First, the organization must have the digital skills to be able to understand what is really important for implementing the top-priority use cases. Providing appropriate digital skills at all layers of the organization is critical for understanding not only the opportunity, but the risks and limitations.

Second, when it comes to adopting the perquisite technologies, companies should take a platform approach, making sure to an appropriate platform with the right guardrails as well as the right governance and organizational frameworks. This is critical for the rapid scaling of the use cases.

Third, companies need build a multidimensional ecosystem, with various technology and data partners. Frequently companies have an abundance of data that are fairly heterogeneous. To ensure the data diversity needed for clinical trials, companies need to be able to access large amounts of data across countries. This makes it critical to forge partnerships with two types of third-party data vendors – those that sell the software and those that implement it.

Learn more about BCG's presence during J.P. Morgan Healthcare Conference here.