Every breath shifts a lung tumor by millimeters to centimeters, making precision radiation therapy especially challenging. Six years ago, Mohamed Abazeed, MD, PhD, and his team, then at Cleveland Clinic and now at Northwestern University, were in the early stages of an ambitious idea: could artificial intelligence (AI) be used to personalize lung cancer treatment and improve outcomes for patients? Today, this vision is becoming reality with promising clinical studies and new tools, which have the potential to redefine precision radiation therapy.
Dr. Abazeed and his team’s latest study in Nature’s Precision Oncology journal introduces iSeg (individualized segmentation), a new AI technology designed to help physicians more accurately identify lung tumors and optimize radiation treatments. For Dr. Abazeed, this milestone represents both innovation and continuity of purpose.
“Our program has always been about transforming lung cancer data into tools that deliver actionable insights to physicians so they can better serve each patient,” explained Dr. Abazeed. “The aim is to ensure treatment is individualized, precise, and effective.”
From Prediction to Personalization
The journey began in 2018 with an image-based AI platform called iGray. By analyzing patient scans, iGray identified which tumors were most likely to recur and, critically, which radiation doses were most effective.
“The tool didn’t just predict outcomes—it provided physicians with the dose guidance in order to prevent treatment failures,” said Dr. Abazeed, who relocated his lab to Northwestern University in 2020, and currently serves as Professor and Chair of the Department of Radiation Oncology. In 2022, the technology advanced to a clinical trial, marking the first time in cancer care that an AI system was formally advising physicians on how to deliver radiation. Physicians remained in control but were asked to document why they overrode the system, ensuring accountability and real-world integration.
The trial has quickly become one of the fastest-accruing studies at Northwestern’s Lurie Cancer Center, with 50 of 70 patients already enrolled. “It’s a pivotal, first-in-class study,” Dr. Abazeed noted, “demonstrating that AI-guided care is not only possible but already taking root.’”
Introducing iSeg: A Sharper Lens on Tumors
Building on this success, the team turned to a new challenge: tumor segmentation. Put simply, if physicians cannot precisely define a tumor’s boundaries, even the most carefully calibrated radiation doses may miss the mark.
“Physicians face a difficult balance,” said Dr. Abazeed. “They may over-contour, irradiating more tissue than necessary to avoid missing the tumor, or under-contour, failing to capture the full extent of disease. Both scenarios reveal the challenge of achieving true precision in radiation therapy.”
Enter iSeg. By distinguishing thousands of grayscale levels on CT scans, far beyond human vision, iSeg can precisely identify tumors while accounting for lung movement during breathing. Early results are striking: the tool reduces unnecessary radiation volume by approximately 30%.
“It’s about shaping care to match the needs of each patient,” Dr. Abazeed explained. “Not a standardized template, but care defined by true precision.”
Expanding the Suite: iTox and Beyond
The Northwestern team is not stopping at radiation dose guidance and tumor segmentation. Another tool in development, iTox, aims to predict which patients are most likely to suffer from radiation-induced pneumonitis, a dangerous side effect.
“The goal is a suite of individualized tools that seamlessly integrate into clinical workflows to make radiation oncology faster, safer, and more effective,” said Dr. Abazeed.
He stressed accessibility: “This isn’t just for big academic centers. Our goal is to democratize these tools so that patients in community hospitals and rural areas also benefit. AI can help every physician, regardless of experience level, deliver higher-quality care.”
From Bench to Bedside—Faster Than Expected
In medicine, it often takes decades for research to translate into clinical practice. Here, transformative progress has occurred in just six years. “What once spanned a medical generation is now unfolding within a few trips around the sun,” said Dr. Abazeed. He credits early support from the American Lung Association as catalytic, providing critical support at a pivotal time in the research.
“We’re deeply grateful,” concluded Dr. Abazeed. “That early funding validated our vision. Even when the funded project wasn’t the exact one pursued, the momentum it created was critical, demonstrating the vital role of supporting early- and mid-career investigators.”
For more information about the American Lung Association Research Institute and our commitment to lung cancer research, visit: Lung.org/research.
Blog last updated: November 3, 2025
