Artificial intelligence could play a pivotal role in expanding access to radiotherapy, one of the most essential cancer treatments worldwide. New findings from the international ARCHERY trial show that AI-driven software can deliver high-quality radiotherapy planning for cervical and prostate cancer, potentially helping to close critical care gaps, particularly in low- and middle-income countries.
Cervical cancer remains a major global health challenge, with around 94% of deaths occurring in lower-resource settings. In 2022 alone, approximately 350,000 women died from the disease. Despite radiotherapy being the primary curative treatment, access remains severely limited: only about 10% of patients in low-income countries receive it, largely due to shortages of trained specialists.
Large-scale trial
The ARCHERY trial, led by researchers from University College London and London School of Hygiene & Tropical Medicine, evaluated the effectiveness of AI-assisted radiotherapy planning across hospitals in India, South Africa, Jordan and Malaysia. The study included more than 1,000 patients with cervical, prostate and head and neck cancers.
Results presented at ESTRO 2026 showed that the AI system achieved a high standard of radiotherapy planning in over 95% of cervical cancer cases. For prostate cancer, the figure reached 85%, which researchers say is still sufficient for routine clinical use. Results for head and neck cancers are expected later this year.
Unlike many earlier AI studies, which are often small and conducted in single centres, this trial provides robust, multi-country evidence supporting real-world implementation.
From weeks to hours
Radiotherapy planning is a complex and time-intensive process. Clinicians must carefully outline tumours and surrounding tissues on CT scans, while medical physicists determine optimal radiation beam configurations. This workflow can take several hours spread over days or even weeks, depending on staff availability.
The AI software automates these steps, identifying target structures and calculating optimal radiation delivery in just over an hour. According to lead investigator Ajay Aggarwal, this reduction in planning time could significantly shorten waiting lists and increase treatment capacity.
By streamlining workflows, the technology may help healthcare systems treat more patients without requiring a proportional increase in specialised staff. A key advantage in regions facing workforce shortages.
Substantial implications
The implications for global health are substantial. Radiotherapy contributes to the cure of around 40% of cancer cases, yet millions of patients worldwide still lack access. AI-enabled planning could help scale up services more efficiently and support initiatives such as the World Health Organization’s cervical cancer elimination strategy.
Co-investigator Mahesh Parmar emphasised the importance of rigorous evaluation. Large-scale trials like ARCHERY, he noted, are essential to demonstrate that AI solutions are safe, effective and applicable across diverse healthcare settings.
External experts also see strong potential. Matthias Guckenberger of University Hospital Zurich highlighted that AI-supported plans were reviewed and validated by clinicians in most cases, confirming their clinical reliability.
Efficiency gains
While the technology is particularly valuable in resource-constrained environments, its benefits are not limited to them. Even in high-income countries, AI could improve efficiency, reduce costs and free up specialist time for more complex cases.
By accelerating planning and optimising resource use, AI-driven radiotherapy tools may help healthcare systems worldwide deliver faster, more equitable cancer care. As further results emerge and implementation expands, such innovations could become a cornerstone in the global fight against cancer.
AI radiotherapy
Last year we reported on the fact that AI is improving radiotherapy for oral cavity cancer by making treatment planning faster and more precise at Leiden University Medical Center. AI supports clinicians in automatically outlining organs on CT scans and generating radiation plans based on large datasets of previous treatments.
This reduces planning time significantly, from up to eight hours to just one or two, while allowing clinicians to review and refine AI-generated proposals. The technology also improves dose distribution, helping to better protect healthy tissues such as swallowing muscles and salivary glands.
As a result, patients experience fewer side effects, including swallowing difficulties, and benefit from improved quality of life. AI additionally enables adaptive radiotherapy, where treatment plans can be adjusted during therapy based on daily imaging, further enhancing precision and effectiveness.
