The IAEA's Director General travelled to a hospital in Mombasa to inaugurate a new linac used for life saving cancer care, provided under the IAEA's Rays of Hope initiative.
The IAEA's global radiotherapy database, the Directory of Radiotherapy Centres (DIRAC), has integrated artificial intelligence (AI) into its workflow to improve data collection and better support research on access to cancer care.
More than half of all cancer patients require radiotherapy at some point during their treatment, yet access to this life-saving service remains unequal. While most high-income countries have enough machines to provide timely care, availability drops sharply in less-resourced settings, with some countries having no radiotherapy services at all.
To enable effective investment to close this gap, reliable information is needed on where treatment is available, which equipment is operational and when it was last verified. This information helps governments, healthcare providers and development partners make informed decisions.
Supporting Better Cancer Care with Data
The IAEA's Directory of Radiotherapy Centres helps experts assess radiotherapy infrastructure, plan radiation oncology services, benchmark resources and support research on equitable access to cancer care. The database includes nearly 8700 radiotherapy centres, over 17 000 teletherapy machines and more than 3000 brachytherapy units worldwide.
Researchers from 114 countries have used DIRAC data in more than 300 scientific publications, including over 200 in the past five years. It also informed the 2024 Lancet Oncology Commission on Radiotherapy and Theranostics and more than 220 other studies in health economics.
International organizations also rely on DIRAC data to analyse access to cancer technologies.
The World Intellectual Property Organization (WIPO), for example, has incorporated DIRAC data into its Global Innovation Index to assess the adoption and availability of radiotherapy technology worldwide. It will use updated data in its next report due in September 2026.
"The WIPO-IAEA collaboration has shown that radiotherapy equipment data can be a valuable measure of technology adoption in health," said Sacha Wunsch-Vincent, Head of Section in WIPO's Department for Economics and Data Analytics and co-editor of the Global Innovation Index. "Making the update process more systematic can strengthen that global evidence base and support more informed policy discussions."
DIRAC's data on equipment availability highlights gaps in access. In high-income countries, one teletherapy machine serves on average about 140 000 people, while in low-income countries this figure rises to over 15 million.
To meet the benchmark of one machine for every 500 patients needing radiotherapy, low-income countries would need more than 21 times their current number of machines. Overall, only about 25 per cent of countries met minimum radiotherapy resource requirements in 2024.
AI Supporting More Accurate and Timely Data Updates
Because radiotherapy networks evolve, with new centres opening, equipment being replaced and services expanding, keeping records up to date requires continuous data collection.
"DIRAC has long helped countries plan cancer services and understand gaps in access. As changes in radiotherapy resources accelerate, we need update processes that are faster, more robust and still firmly grounded in expert validation," said Mauro Carrara, Head of the IAEA's Dosimetry and Medical Radiation Physics Section.
To support this, the IAEA has introduced a new workflow that complements traditional data collection as part of its broader efforts to apply advanced technologies, including artificial intelligence, to strengthen its programmatic activities.
The new process uses AI to identify publicly available information on radiotherapy centres and equipment and convert it into structured records. These records are reviewed and validated by experts before being incorporated into DIRAC.
This approach prioritizes data quality and reliability. Human oversight remains central at every stage, ensuring that faster updates do not come at the expense of accuracy. Pilot studies in several countries have demonstrated the potential of this approach to improve data precision, expand geographic coverage and shorten update cycles.
"Every reliable data point in DIRAC can increase the accuracy and impact of health economic analyses as well as help decision-makers, including potential investors and donors, better understand where cancer treatment capacity exists and where gaps remain," said May Abdel-Wahab, Director of the IAEA's Division of Human Health. "By strengthening data systems with AI, while still maintaining rigorous expert validation, we can ensure that countries receive timely, high-quality information that ultimately supports better cancer care for patients in greatest need."
