An Automatic Needle and Seed Planning Algorithm for Head and Neck Seed Implant Brachytherapy
Z Xiao1,F Zhou1,2, B Liu1,2*, Z Ji3, H Sun3, Y Jiang3, J Wang3, Q Wu4, (1) Image Processing Center, Beihang University, Beijing, CN (2) Beijing Advanced Innovation Center for Biomedical Engineering, Beihang University, Beijing, CN (3) Department of Radiation Oncology, Peking University Third Hospital, Beijing, CN (4) Duke University Medical Center, Durham, NC
Presentations
SU-H300-IePD-F7-3 (Sunday, 7/10/2022) 3:00 PM - 3:30 PM [Eastern Time (GMT-4)]
Exhibit Hall | Forum 7
Purpose: Seed implant brachytherapy (SIBT) is an effective treatment modality for head and neck (H&N) cancers. Due to the complex anatomy with many critical organs (such as bones and blood vessels) to avoid, current clinical planning requires manually placing needle paths which is time-consuming. The purpose of this work is to develop and validate an automatic planning method for H&N SIBT.
Methods: A novel strategy inspired by clinical experience was proposed to automatically setup potential needle paths. Firstly, a reference entry plane was determined according to target position and a series of non-coplanar entry planes were generated by step-wisely rotating the reference plane around the target. Then, based on a concept of target coverage ratio, entry planes with the largest target exposure were chosen from which candidate needle paths were generated with scrutiny to avoid needle collisions and punctuating the bone and critical organs. Afterwards, a greedy heuristic-based optimization method based on dose-volume constraints was developed to optimize the seed positions. A cohort of 25 patients were collected and the generated plans were compared with clinical plans.
Results: The method could automatically generate clinically acceptable plans in 0.2±0.2 minutes, which meet the requirement of real time clinical operation. The resulting V₁₀₀, V₁₅₀, V₂₀₀, D₉₀ and D₁₀₀ for targets were 96.8±1.7%, 78.5±11.2%, 55.1±17.1%, 130.9±14.6% and 64.3±16.7% which were comparable to clinical plans used for treatment (96.5±3.7%, p=0.71; 79.5±13.6%, p=0.20; 55.3±21.3%, p=0.96; 131.7±21.7%, p=0.91; 66.2±18.6%, p=0.44). The generated needles effectively avoided the critical organs, and the numbers of the used needles (10.2±2.7) and seeds (32.7±10.7) were similar with clinical plans (11.4±3.7, p=0.74; 32.7±11.5, p=0.12).
Conclusion: The proposed method can generate clinically acceptable plans quickly, and has a good potential for clinical usage.
Funding Support, Disclosures, and Conflict of Interest: This work was supported by the National Key Research and Development Program of China under Grant 2019YFB1311300 and 2019YFB1311301.
Keywords
Brachytherapy, Treatment Planning
Taxonomy
TH- Brachytherapy: Treatment planning using machine learning/automation
Contact Email