One-Click Breast Cancer Treatment: AI Automates Radiation Planning, Saving Time and Improving Care
"New AI framework slashes breast cancer radiation treatment planning time by 50% while maintaining the quality of care."
Radiation therapy is a cornerstone of breast cancer treatment, particularly after breast-conserving surgery. Traditional methods, however, can expose the entire breast to high doses of radiation, leading to potential complications. Achieving uniform radiation dosage in the breast, given its complex shape, has always been a challenge.
To overcome these limitations, advanced techniques like 3D image-based treatment planning, conformal radiation therapy (CRT), and intensity-modulated radiation therapy (IMRT) have been developed. These approaches aim to deliver precise radiation doses to the tumor while minimizing exposure to surrounding healthy tissues. However, these methods can be time-consuming and require specialized expertise.
Now, a groundbreaking solution is emerging: an automated field-in-field (FIF) planning framework powered by artificial intelligence (AI). This innovative approach promises to streamline the radiation treatment planning process, making it faster, more efficient, and more accessible to healthcare providers.
Automation in Breast Radiation Planning
Recent studies have evaluated automated planning for whole-breast irradiation using systems such as mCycle integrated with the Monaco Treatment Planning System. The 2025 study compared step-and-shoot intensity-modulated radiation therapy (IMRT) and volumetric modulated arc therapy (VMAT), focusing on whether automation could standardize plan quality across a diverse patient cohort. Earlier research developed an automatic whole-breast radiation planning system based on two intensity-modulated tangential fields to enable near-real-time planning. Other work has evaluated machine learning-based automated planning in clinical breast radiation treatment and developed algorithms using anatomical characteristics from 400 breast cancer patients.
From Manual Planning to Automation
Breast radiation planning has traditionally involved manual clinical planning, including techniques such as helical tomotherapy and hybrid IMRT. A study of 25 patients with locally advanced breast cancer compared automatically generated volumetric modulated arc therapy plans with manually created clinical helical tomotherapy plans. Another study retrospectively planned 100 clinical breast cancer cases with a single-click automated planning code that automated plan generation, beam generation, gantry and collimator angles, open segments, dynamic IMRT fluence, and calculations. These studies illustrate efforts to reduce the number of manual planning steps, while comparisons with established clinical plans remain necessary.
The Evolution of Automated Planning
Radiation treatment planning has progressed substantially over the past several decades through advances in hardware and algorithms. These developments helped establish modalities including intensity-modulated radiation therapy and volume modulated arc therapy, which the cited review associates with improved patient outcomes and quality of life. More recent work demonstrated the feasibility of an automated treatment planning portfolio for whole-breast radiation therapy that can encompass different treatment paradigms and clinical preferences. Together, these milestones show a progression from advanced radiation modalities toward broader automation of planning workflows.
How Does Automated FIF Planning Work?
The automated FIF framework combines the Eclipse Scripting application programming interfaces (APIs) with user-executed programming in Windows. This integration allows for a one-click option within the treatment planning system (TPS) to automate breast FIF planning.
- Scripting APIs: Designed to automate clinical workflows in radiation oncology.
- User-Executed Programming: Integrates the scripting application with user-defined programs to automate the field-in-field technique.
- Eclipse Scripting Applications: Provide an interface between the treatment planning system server and client, enabling the execution of programs that create dose clouds and adjust planning parameters like multi-leaf collimator (MLC) placements and monitor unit values.
- One-Click Option: Allows all tasks to be performed with a single click within the treatment planning system, including automated pushback of the proposed plan.
Automated Planning Across the Workflow
Recent research describes automation across multiple stages of whole-breast irradiation planning. These stages include segmenting treatment targets and normal structures, selecting beam angles, optimizing wedge angles for tangent beams, optimizing segment shapes and weights, and creating inverse-planned IMRT plans. The cited review states that automation can reduce planning time while producing comparable or better plan quality. Related literature also notes that automated solutions for adding subfields exist for breast plans, while a separate study reported planning time falling from 33 minutes to 5 minutes in canine and feline nasal tumor planning, a different clinical application.
Validation Still Matters
Automation does not remove the need to evaluate treatment plans against manually produced clinical plans. One study tested an automated field-in-field framework for breast forward IMRT on 40 retrospective breast cancer data sets. It compared the automated plans with manual plans using mean and maximum target dose, homogeneity index, and the volume receiving the prescription dose at the 100% or 102% isodose line after normalization to the manual plan's mean target dose. This evaluation approach highlights that automated planning must be assessed with dosimetric measures rather than assumed to be equivalent to manual planning.
The Future of Breast Cancer Treatment Planning
The combination of scripting applications and user-executed programming for automated breast field-in-field planning represents a significant advancement in planning efficiency without compromising plan quality. As AI and automation continue to evolve, we can expect even more sophisticated solutions that improve cancer treatment outcomes and enhance the quality of life for patients.
Toward Personalized Automation
Evidence points toward automated breast radiation planning that accounts for patient anatomy and clinical context. One study developed and validated automated IMRT planning algorithms using anatomical characteristics in 400 breast cancer patients without lymph node involvement. A 2024 study identified a limitation in existing automated methods: they lacked a comprehensive, end-to-end solution able to meet varying standards of care. Its proposed portfolio was designed for intact breasts and tailored to individual patient factors, clinical approaches, and available resources.
AI-Assisted Planning Horizons
Deep learning is a major future direction for radiation treatment planning, particularly for automatic segmentation. A 2024 review describes the rapid development of artificial intelligence techniques as promising for radiation oncology. It identifies radiotherapy planning as a strong candidate for automation because the process is largely computational and closely connected to medical imaging. Progress in this area could expand AI involvement beyond plan optimization to additional imaging-based planning tasks.
Supporting Consistent Clinical Care
The mCycle automated planning system has been integrated into the Monaco Treatment Planning System for whole-breast irradiation using step-and-shoot IMRT and VMAT. The cited study evaluated whether this automation could standardize plan quality across a diverse patient cohort while comparing the dosimetric characteristics of the two treatment techniques. This suggests that the practical impact of automation is not only faster planning, but also more consistent evaluation of treatment options across patients. Clinical teams remain central to interpreting those comparisons and deciding how plans should be used in care.