Dr. Todd McNuttA Tool for Obtaining the Best Radiation Therapy Treatment Plans
In the pursuit of better cancer treatments, Dr. Todd McNutt and his team at Johns Hopkins have developed Oncospace, a groundbreaking data-mining system that leverages artificial intelligence to enhance radiation therapy planning. By analyzing the experiences of over 5,000 patients and integrating various medical factors, Oncospace helps oncologists craft optimal treatment plans tailored to individual needs while minimizing side effects. With FDA approval for multiple cancer types and a recent acquisition by Sun Nuclear, this innovative tool—now rebranded as Plan AI—promises to elevate the standard of care for patients globally, ensuring that even smaller practices can deliver advanced, evidence-based radiation therapy that matches the quality of prestigious institutions.
Advances in cancer medicine don’t always come in the form of a new drug or therapy. After years of seeing valuable information collected in cancer care left untapped, radiation oncology physicist, Todd McNutt, Ph.D., envisioned a data-mining system that could learn from every patient’s prior experience and improve treatment for the next.
The vision began more than a decade ago, when Dr. McNutt partnered with head and neck cancer radiation oncologist Harry Quon, M.D., and computer scientists Russell Taylor, Ph.D., and Misha Kazhdan, Ph.D. Together, they built a system that could mine data from thousands of patients treated with radiation therapy.
By connecting diagnoses, treatments, side effects and outcomes, the tool, called Oncospace, helped guide therapy, evaluating the treatments that worked best for a particular cancer and learning from those that resulted in less-than-favorable outcomes to generate an optimal treatment plan.
Oncospace simultaneously took into account and connected variables including age, underlying health conditions and other treatments patients were receiving to determine how these factors influenced toxicities and response to treatment. As a result, Oncospace sharpened treatment plans, lowered toxicities, and improved quality of life.
Every patient who receives treatment adds to the data, so it is ever improving as new information — and more types of data, such as structured physician notes and imaging — are included.
It was a tool that no other cancer center in the world had, so Dr. McNutt took steps to make it available beyond Johns Hopkins.
Better, Smarter, Faster
In 2019, with support from grants, donors, and Johns Hopkins Technology Ventures, Dr. McNutt, programmer Michael Bowers, and colleagues, including Pranav Laksminarayanan and Julie Shade, and a few outside partners, founded Oncospace, Inc. The current version is boosted by AI to even further improve the quality of radiation therapy treatment plans.
“We built a tool that can predict — for each patient — the best achievable treatment plan before the planning process even begins,” says Dr. McNutt. “We think it will raise the level of quality for everyone, so a small practice can use the tool and obtain the same standards and level of quality as an academic research institution like the Kimmel Cancer Center, and that benefits patients everywhere.”
A Smarter Way to Plan Care
Drawing on a knowledge base built from the histories of over 5,000 patients and growing, Oncospace gives radiation oncologists and dosimetrists an AI “assistant” — a powerful starting point optimized to balance tumor control with protecting nearby healthy tissue. As Dr. McNutt and colleagues refine their plans, Oncospace is a guide always pointing toward the most optimal plan and also indicating when the plan quality is less than adequate.
“Unlike other commercial systems, which require hospitals to build their own models from scratch, a time-consuming process that often draws on only small patient cohorts, Oncospace offers an advanced, validated model from day one,” Dr. McNutt says.
It was designed to raise the level of care anywhere it is used, providing consistency, built-in safety checks, and the ability to compare results.
Global Reach
Oncospace has gone through rigorous studies and has received Food and Drug Administration approval for multiple cancer types, including prostate, gynecological, head and neck, liver, pancreatic, lung and esophageal cancers.
In April 2025, the company was acquired by Sun Nuclear (Mirion Medical), a global leader in radiation oncology quality assurance. With Sun Nuclear’s worldwide reach, Oncospace, which was renamed Plan AI, will expand from two development sites — Johns Hopkins and Montefiore Hospital — extending the benefits of the Kimmel Cancer Center-born innovation to patients around the globe.