Automated AI planning technology may generate radiotherapy treatment plans in under one second, potentially reducing patient wait times from weeks to days, though implementation requires addressing workforce shortages, infrastructure gaps, and payment system incentives that currently delay treatment.
- Deep-learning systems like AIRT can generate single-arc prostate radiotherapy plans in under one second using GPU processing, matching or exceeding traditional planning methods.
- Current radiotherapy workflows create delays of 10-14 days between consultation and treatment start, with some regions reporting waits exceeding 477 days for cervical cancer care.
- Automation alone cannot solve access problems; solutions require workforce development, appropriate infrastructure procurement, equipment maintenance, and training matched to patient demand.
- Patient costs extend beyond treatment charges to transportation, accommodation, and food, making geographic access and treatment speed critical economic factors in cancer care equity.
Research promising one-second treatment planning prompted discussion of workforce shortages and delays in cancer care at the IAEA Scientific Forum in Vienna.
“Technology has to bridge the gap,” said Deepak Khuntia, senior vice president of medical affairs and chief medical officer at Varian, a Siemens Healthineers company. Speaking in Vienna, he argued that existing medical education capacity could not produce enough clinicians to meet growing demand. He presented automation as a potential response to capacity constraints, alongside efforts to develop the workforce.
One-second planning: the claim
“From the time that planning scan is done, within one second, the treatment plan is completed,” he said. His description encompassed segmentation of normal structures and targets, optimization, and plan quality assurance.
“This is a research project,” Khuntia emphasized, adding: “Be careful with how you interpret this.” He said the publication is planed for later that year.
Research context and limits
IAEA Scientific Forum Vienna 2026
A separate July 2026 preprint describes a foundation-model agent performing daily cone-beam CT (CBCT)-guided adaptive planning in under two minutes. Evaluations covered head-and-neck, lung, abdominal, and prostate cancers, including photon and proton therapy. The authors explicitly describe clinician intervention and final approval within a human-in-the-loop framework.
Where time accumulates
“It can be 10 days to two weeks from the consult before the patient actually got treated,” Khuntia said of the workflow he described. His example included several days before the planning CT, followed by work by the clinical and planning team. He did not present this interval as a measured national average.
Thomas Pascual of the Philippine Nuclear Research Institute linked waiting time to workforce, physical infrastructure, and equipment. His discussion covered consultation through imaging, biopsy, and treatment, with responses including appropriate procurement, maintenance arrangements, training, and facilities matched to patient demand.
IAEA Scientific Forum Vienna 2026
“You get paid less if you do all of the planning on the same day as the consult,” Khuntia said, arguing that U.S. reimbursement can encourage separating steps across days.
“That’s not good medicine, but that’s how you make medicine profitable in the U.S.”
The human cost
“These types of delays are unacceptable,” Khuntia said, recounting an African governor’s report of a 477-day average wait for cervical-cancer radiotherapy in his state.
“How does she get to the nearest center that offers radiation?” asked Zainab Shinkafi-Bagudu, CEO of the Medicaid Cancer Foundation and president-elect of the Union for International Cancer Control.
She pointed to transportation, accommodation, and food costs, arguing that patients’ lived experience should inform the economic case for cancer investment. Her intervention extended the discussion beyond treatment charges to the practical costs of reaching and remaining near a center.
