S3005
Invited Speaker
ESTRO 2026
5330 Brachytherapy workflow optimisation in breast cancer Jean-Michel Hannoun-Levi Radiation Oncology, Antoine Lacassagne Cancer Center, Nice, France How workflow optimization can significantly improve the efficiency of breast brachytherapy while maintaining high standards of quality and safety? This is the subject we are going to talk about. Workflow optimization is defined as the process of streamlining each step of patient management to reduce delays, resource burden, and unnecessary complexity. In breast brachytherapy, several bottlenecks are identified along the care pathway, including pre-implant consultation, imaging (pre- and post-implant CT scans), the implant procedure itself, treatment delivery, and post-implant follow-up. These steps, when poorly coordinated, can lead to prolonged treatment times and inefficiencies in patient flow. The consequences of these inefficiencies are multidimensional. From the patient perspective, delays and complexity may reduce comfort and satisfaction. Organizationally, workflow bottlenecks strain resources and limit throughput. They may also affect treatment quality and safety, increase workload for healthcare professionals, and generate additional economic costs. Brachytherapy inherently offers key advantages by delivering a high radiation dose to a small target volume over a short period making it an ideal modality for optimization. The challenge is therefore not technological capability, but rather organizational efficiency. The presentation highlights several strategies to address these issues. First, simplification and integration of workflow steps can reduce redundancy. Second, technological innovations, including imaging improvements and artificial intelligence, can accelerate planning and execution. Third, adopting ultra-short treatment schedules, such as very accelerated partial breast irradiation, significantly shortens overall treatment time. A major example discussed is single-fraction partial breast irradiation (sfPBI), currently being evaluated in a multicenter phase II trial for low-risk invasive and in situ breast carcinoma. This approach represents the ultimate form of workflow optimization by condensing treatment into a single session, thereby minimizing patient burden and maximizing resource efficiency. In conclusion, optimizing brachytherapy workflows is essential to fully exploit the clinical benefits of this technique. By combining organizational improvements, technological support, and innovative treatment regimens, it is possible to enhance patient
experience, improve efficiency, and maintain high- quality oncologic outcomes.
5331 Brachytherapy workflow optimisation in prostate cancer Justinas Jonu š as Radiation Oncology Center, National Cancer Center, Vilnius, Lithuania Brachytherapy remains a cornerstone of curative radiotherapy for prostate cancer. However, its workflow remains labor and resource-consuming. Additionally, it is heavily dependent on the expertise of the physician and constrained by intraoperative time pressure as the patient is under general anesthesia. As the field seeks to improve its workflow through various optimization strategies, a growing body of evidence points to hardware and software innovations that can meaningfully streamline it. First, AI-driven automation of time-critical intraoperative steps offers the most immediate gains. AI-based needle reconstruction has been prospectively validated in ultrasound-guided HDR prostate brachytherapy, reducing reconstruction time by more than 15 minutes while maintaining accuracy, with only 4.5% of needles requiring manual correction 1 . Machine-learning-based treatment planning generates LDR brachytherapy plans in under one minute that match those created by experienced medical physicists 2 . At the same time, deep convolutional neural networks predict dose distributions 300 times faster than conventional simulations, with dosimetric differences of less than 2% in prostate cases 3 . Additionally, automated segmentation of targets and organs at risk is the single largest category of AI research, driven by the need to reduce inter-observer variability and accelerate planning 4 . Second, next-generation dose optimization moves beyond conventional DVH evaluation. Data-driven spatial modeling reduces hot-spot volumes by 29% using convex estimators that capture high-dose regions 5 . It is information that is usually missed in the standard DVH metrics. GPU-accelerated Monte Carlo inverse planning improves dose homogeneity and lowers urethral dose within intraoperative time constraints 6 . Novel source delivery capabilities, including rotating shield systems and simultaneous thermo-brachytherapy applicators, achieve 20–31% reductions in organ-at-risk dose while maintaining target coverage. However, some issues need to be solved before the transition to clinical use. Firstly, most AI models are trained on single-center cohorts, which limits generalisability. Secondly, prospective multicentric validation remains scarce. Moreover, anatomic
Made with FlippingBook - Share PDF online