AI Support in the Patient Care Journey

Northwell Health Develops a Fast Track for Cancer Diagnoses

In 2023, Northwell Health introduced iNAV, an AI tool developed by three physicians to analyze MRIs and CT scans for cancer. iNAV looks at scans that patients obtain for other health reasons, presenting a chance to detect cancers much earlier. When the tool flags an issue, a multidisciplinary team reviews the findings, enabling care navigators to expedite biopsies.

iNAV quickly proved its value, reducing biopsy wait times from 22 days to just seven and treatment time from 56 days to 34. At scale, such improvements can be momentous for systems such as Northwell, which has 21 hospitals throughout the state of New York.

“The sooner patients get in, the sooner they can start treatment, and the better the outcomes are,” says Dr. David Rivadeneira, director of the Northwell Health Cancer Institute at Huntington and physician in chief for clinical strategic initiatives for Northwell Health’s eastern region.

Initially piloted for pancreatic cancer, iNAV now detects other cancers as well. Northwell Health also uses AI for treatment planning and clinical trial matching, synthesizing massive amounts of data to help clinicians optimize and personalize care.

“Everybody’s tumor is different, and everybody’s cancer responds in a different way,” Rivadeneira says.

Time named iNAV one of the best inventions of 2024, recognizing early detection’s critical role in better outcomes.

“We’re very excited about this platform,” Rivadeneira says. “I think many health centers will look to adopt something like this.”

EXPLORE: AI in customer experience enhances the patient's journey.

MUSC Works Toward an Agile Approach to AI

In Charleston, S.C., the Medical University of South Carolina Health has leveraged an AI platform for a growing number of use cases. MUSC initially deployed the solution to streamline patient registrations before branching out to care-gap outreach and insurance authorizations.

Enterprise Chief Digital Transformation Officer Crystal Broj says this gradual expansion reflects an important aspect of AI: “You have to be willing to be iterative.”

When MUSC added a copay feature to patient registrations, it didn’t work flawlessly at first. But by adding features slowly, the health system could incorporate patient feedback and refine the platform before rolling it out to its 700 offices statewide.

Being agile also allows for proper change management, Broj adds, ensuring that staffers understand AI’s benefits and have the training to use new tools effectively.

That’s especially important when AI is working at scale. In 2025, the platform sent 1.7 million reminders to patients, which freed office staff to direct their attention to other tasks, she says.

When MUSC saw an opportunity to use AI for care-gap outreach, it started with patients who were overdue for mammograms. The platform’s AI agents reviewed medical records and sent messages to eligible patients with a self-scheduling option.

“We turned on this function on a Thursday night, and by Friday morning, 129 women had scheduled appointments,” Broj says.

MUSC has continued to expand its AI-assisted care-gap outreach for other needs, such as flu shots and well-child visits. “The goal is to make it as easy to work with your health system as it is to order a book on Amazon,” she says.

OSF HealthCare Finds a New Approach to Concussion Care

OSF HealthCare, an integrated health system with hospitals and clinics serving Illinois and Michigan, is testing a new, AI-enabled app that provides a timely, objective measure of whether athletes have concussions.

“It became increasingly clear to me that our ability to identify people with concussions reliably is very poor,” says Dr. Adam Cross, assistant professor of pediatrics at the University of Illinois College of Medicine Peoria and director of the Children’s Innovation Lab at Jump Trading Simulation and Education Center, a collaboration between the university and OSF HealthCare. “There had to be a better way.”

Currently, coaches use time-consuming tools that rely heavily on athletes’ self-reported symptoms, but those can be influenced by a player’s desire to return to the field, Cross says. So, together with co-developers from the Grainger College of Engineering at the University of Illinois Urbana-Champaign, Cross and his team developed FlightPath.

The percentage of healthcare organizations that have partnered with third-party vendors to develop custom, generative AI solutions

Source: McKinsey, “Generative AI in Healthcare: Adoption Trends and What’s Next,” July 2024

The mixed-reality app uses a gaming approach and asks users to position a bubble around a virtual hummingbird flying within a 3D space on an Apple iPad device. Cross says AI plays several roles in the app, including collecting hundreds of thousands to millions of data points in each two-minute session.

“We use all this data to understand how a person moves within their space and in response to the movements of the bird,” he says. “And we have the hummingbird move in very specific but also semi-randomized ways so that we can elicit what we know is a neurologically complex or meaningful movement.”

OSF HealthCare has partnered with athletes and coaches at Illinois State University, Illinois Wesleyan University and Bradley University to test the solution for two years. This will allow Cross’s team to gather data and compare the app’s performance to other diagnostic tools. “We’ve had a lot of very positive feedback,” he adds.

Rapid advances in AI have alleviated many of the technological obstacles that slowed down early use cases. Today, Cross says, “the bigger issue is knowing how to apply all of this horsepower to meaningful problems and then using them in clinical practice appropriately.”

UP NEXT: Overcome AI implementation hurdles in healthcare.