ESTRO 2026 Research Highlights: Advancing Quality, Efficiency, and Access Through Automation

The 2026 ESTRO Annual Meeting in Stockholm showcased a wide range of innovative research exploring new applications of Radformation solutions spanning adaptive radiotherapy assessment, automated treatment planning, patient-specific QA, and workflow optimization in resource-constrained environments.

Together, these studies highlight a common theme: automation is increasingly enabling clinicians to make more informed decisions, standardize treatment quality, and expand access to high-quality radiotherapy care.

Supporting Treatment Adaptation Decision-Making in Head and Neck Radiotherapy
In the first known publication to explore the use of the recently released ChartCheck Adaptive, David Barbee of NYU Langone investigated the role of automation in offline adaptive assessment for head and neck radiotherapy.

In a retrospective analysis of 90 VMAT H&N patients encompassing approximately 2,800 treatment fractions, Barbee evaluated anatomical and dosimetric changes occurring over the course of treatment. The study leveraged ChartCheck Adaptive to generate synthetic CT datasets using daily CBCT imaging, calculate delivered dose using RadMonteCarlo, and perform comprehensive dose constraint and structure volume analyses.

Data from Barbee's 2026 ESTRO abstract demonstrates systematic volume changes over the course of treatment for parotids and target volumes. Volumes for structures not expected to change show stable volumetric calculations, validating the use of deformable outputs for trend monitoring.

The data collected demonstrated consistent trends in target and organ-at-risk volume changes across a diverse head and neck patient population—notably in PTV and parotid OAR volumes—suggesting that these anatomical changes are not isolated events but instead characterize shifts across all H&N patients sampled. By quantifying when and how often volumes change or target coverage losses occur, the platform provides clinicians with objective data to help identify patients who may benefit from adaptive intervention before clinically significant deviations occur.

Expanding Automated Planning Workflows with Monaco Treatment Planning System

Several presentations highlighted the growing role of EZFluence in streamlining treatment planning workflows, including new evidence supporting its use within the Monaco Treatment Planning System as well as in low-resource settings.

Investigating Plan Quality
Researchers from Fondazione IRCCS San Gerardo dei Tintori presented the first clinical evaluation of EZFluence automated forward planning for breast and whole-brain radiotherapy in Monaco treatment planning system (TPS). The study assessed both the clinical feasibility and workflow integration of automated planning for these common treatment sites.

The team found that EZFluence-generated breast plans demonstrated improved target coverage while reducing high-dose hotspots. Whole-brain plans similarly achieved significant hotspot reductions while maintaining comparable target coverage and reducing monitor units. These findings suggest that automated forward planning can enhance plan quality while supporting efficient clinical workflows across multiple treatment sites within Monaco TPS.

Evaluating Novel Breast Immobilization Approaches
A second study from the same institution investigated a novel immobilization bra designed for patients with larger breast volumes.

Again, using EZFluence-generated plans within Monaco TPS, researchers compared treatment plans created with and without the immobilization device. The analysis found modest but clinically meaningful reductions in low-dose exposure to both the ipsilateral lung and heart when the immobilization system was used.

 Three-view volumetric dose heat map showing EZFluence tangent plan coverage to large breast patient volumes with and without immobilzation bra.

While the primary focus of the study was immobilization, automated planning enabled a consistent dosimetric evaluation framework, helping investigators isolate and assess the impact of the positioning approach itself.

These studies demonstrate how automation can extend planning standardization and efficiency across different treatment planning environments, providing clinics with greater flexibility in adopting automated planning approaches.

Improving Access to Radiotherapy in Resource-Constrained Settings
Investigators from the Medserve-LUTH Cancer Centre in Nigeria evaluated EZFluence for whole-brain radiotherapy planning in an environment characterized by workforce shortages, limited equipment availability, and heavy clinical workloads. The study noted that manual planning processes can become significant bottlenecks in such settings, contributing to delays in care delivery.

Consistent with other studies evaluating automated forward planning, EZFluence generated plans with excellent uniformity and conformity while dramatically reducing planning time compared with manual approaches. The authors concluded that automated planning offers a practical strategy to improve workflow efficiency, standardize treatment quality, and help overcome technical and staffing limitations.

These findings underscore the potential for automation not only to improve efficiency in established radiotherapy programs but also to support broader global access to quality cancer care.

Validating Monte Carlo-Based QA and Log File Analysis
Patient safety and treatment accuracy remained another important theme at ESTRO 2026, with new research validating the performance of RadMonteCarlo for patient-specific quality assurance.

Researchers at St. Vincent's Private Hospital evaluated RadMonteCarlo for both pre-treatment QA and log file-based verification across 40 patient plans on an Elekta Versa HD linear accelerator. Calculated doses were compared against absolute dose measurements and 2D array measurements across multiple treatment sites.

Results demonstrated strong agreement between RadMonteCarlo calculations and measured doses, with average differences below 2% across prostate, spine, pelvis, and abdomen cases. Gamma passing rates also showed excellent agreement for both pre-treatment and log file verification analyses.

Beyond the dosimetric performance, investigators reported a substantial workflow impact, estimating that RadMonteCarlo can reduce PSQA workload by approximately 60% compared with traditional measurement-based approaches.

Presentations on the ESTRO Main Stage
In addition to exciting new research, Radformation hosted a series of educational presentations featuring clinical leaders sharing their experiences implementing automation, offline adaptive assessment, and QA solutions to address real-world clinical challenges.

Advancing AI AutoContouring: A Clinical Evaluation of Model Performance at MAASTRO
Recently, Radformation debuted AutoContour v2.7, the latest version of our auto-segmentation tool that combines the best of AutoContour and Limbus Contour into one solution. Hendrik Hansen of MAASTRO presented a methodological evaluation of the combined software, comparing the newly unified platform against their established Limbus Contour workflow. The assessment examined both processing speed and contour quality.

The evaluation demonstrated substantial efficiency gains (as much as 9x) through cloud-based processing. A batch test of patient scans using the cloud processed all patient contours in 24 minutes compared to 3.3 hours with Limbus Contour on-premises hardware. Clinical reviewers also found that the vast majority of evaluated structures were equivalent to or improved compared to legacy models, while highlighting the benefits of AutoContour's expanded structure library, enhanced user interface, and advanced post-processing capabilities.

Closing the Gap Between Planned and Delivered Dose: A Multi-Centre Experience with Automated Offline Adaptive Assessment
Ben Archibald-Heeren outlined the Icon Group’s structured approach to commissioning, implementation, training, and quality assurance for ChartCheck Adaptive, designed to automate quantitative offline adaptive assessment. Highlighting the need for clinical validation of deformable image registration (DIR) and beam modeling, he lauded the major improvements to DIR in ChartCheck Adaptive and discussed the implications of beam model match for TrueBeam and Halcyon machines within the framework of daily dose and volume metrics.

The impact for Icon group has already been substantial: with 20 sites currently installed, they plan to scale to 40+ in the coming months. Since implementation, ChartCheck Adaptive has helped identify clinically significant anatomical and dosimetric changes that prompted more than 100 patient replans in six months, contributing to a substantial increase in adaptive interventions (H&N adaptation rate increased from 6% to 29%) and enabling more data-driven treatment decisions.

Design and Implementation of a Customized Machine QA Program for MR-Guided, Adaptive, and Conventional Linac Systems

Martijn Simons of UMC Amsterdam presented the institution's effort to harmonize machine quality assurance workflows across multiple treatment sites following an organizational merger.

Using RadMachine as a centralized platform, the team unified QA procedures across a diverse fleet of treatment systems. This includes MR-guided technology, for which they closely collaborated with the Radformation team to produce tailored MLC QC and picket fence tests to suit their needs. The presentation highlighted the flexibility of customized QA workflows and demonstrated how a cloud-based approach can improve standardization, efficiency, and oversight while reducing reliance on fragmented processes and manual documentation.

UMC Amsterdam's analysis of Winston-Lutz images displays vector displacement by collimator and gantry angle.

Looking Ahead
The research presented at ESTRO 2026 reflects the expanding role of automation across the radiation oncology ecosystem. From identifying candidates for adaptive intervention and streamlining treatment planning to improving quality assurance and expanding access in underserved regions, these studies demonstrate how automation continues to support both clinical excellence and operational efficiency.

As clinical investigators explore new applications and validate emerging workflows, the evidence presented at ESTRO 2026 reinforces the potential of automation to help clinics deliver high-quality, consistent, and accessible radiotherapy care for more patients around the world.