GE Healthcare & Mass General Brigham Expand Collaboration

Foundation models offer a paradigm shift by providing a reliable and adaptable foundation for developing AI applications tailored to the healthcare sector.

Leading global medtech company, GE Healthcare and Mass General Brigham have expanded their collaboration to integrate medical imaging foundation models into their AI research work.

According to the statement, the expansion will focus on responsible AI practices. Both GE Healthcare and Mass General Brigham have been working to explore AI's potential across various diagnostic and treatment paradigms as part of their decade-long partnership deal.

The integration of medical imaging foundation models into their AI research marks a significant step forward in leveraging AI technology to enhance patient care and outcomes, the statement said.

Dr Keith Dreyer, chief data science officer, Mass General Brigham, underscored the transformative potential of foundation models, emphasizing their ability to complement and enhance existing AI frameworks, such as convolutional neural networks.

This integration, Dr Dreyer noted, holds the promise of making healthcare delivery more efficient, accessible, and equitable for diverse communities.

According to him, the relationship between Mass General Brigham’s commercial AI business (Mass General Brigham AI) and GE HealthCare has helped accelerate the introduction of AI into a range of product offerings and digital health solutions.

‘’With foundation models, we are witnessing the next wave of AI innovation, and it is already reshaping how we build, integrate, and use AI,” Dreyer added.

How Foundation Models Help Develop AI Solutions Tailored for Healthcare?

Traditionally, integrating AI into healthcare systems has entailed the arduous task of retraining models to suit diverse patient populations and hospital settings, leading to increased costs and complexity.

Foundation models offer a paradigm shift by providing a reliable and adaptable foundation for developing AI applications tailored to the healthcare sector.

By streamlining workflow efficiency and improving imaging diagnosis, foundation models have the potential to make healthcare delivery more effective.

Parminder Bhatia, chief AI officer, GE Healthcare, highlighted the collaborative efforts between GE Healthcare and Mass General Brigham in harnessing AI to develop tools that enhance operational effectiveness and productivity in healthcare settings.

"Incorporating responsible AI practices into this phase, we are committed to ensuring these innovations adhere to guidelines, prioritize patient safety and privacy, and promote fairness and transparency across all applications," Bhatia noted.

First AI Application Developed by GE Healthcare-Mass General Brigham Collaboration

In 2017, GE Healthcare, headquartered in Chicago, and Mass General Brigham initiated a 10-year collaboration aimed at exploring the application of AI across diverse diagnostic and treatment methodologies.

Initially, the emphasis of this partnership was on creating applications geared toward enhancing clinician efficiency and improving patient outcomes in diagnostic imaging.

The organizations expressed their intention to evolve this collaboration over time, diversifying into new business models for AI applications in healthcare and expanding product development into additional medical specialties such as molecular pathology, genomics, and population health.

The inaugural AI application resulting from this collaboration is the schedule predictions dashboard within the Radiology Operations Module (ROM), a digital imaging tool designed to optimize scheduling, reduce costs, and alleviate administrative burdens on providers.

This enhancement aims to afford clinicians more time to dedicate to the clinician-patient relationship. The ROM is currently available for implementation in healthcare institutions.

Operational AI-enabled tools play a pivotal role in addressing challenges that can impede patient care, such as healthcare costs and inefficiencies within hospitals.

For instance, the consequences of missed appointments, failure to schedule follow-ups or tardiness, collectively termed missed care opportunities (MCO), can be substantial.

The jointly developed algorithm aims to predict instances of MCO and late arrivals, facilitating better management of urgent, inpatient, or walk-in appointments.

In preliminary testing, the algorithm demonstrated an accuracy rate of up to 96% in correctly predicting missed care opportunities, while maintaining minimal false positives.