ESTRO 2024 - Abstract Book

S395

Brachytherapy - Urology

ESTRO 2024

417

Proffered Paper

Electromagnetic tracking reveals reconstruction errors in intraoperative TRUS – based prostate HDR-BT

Ioannis Androulakis, Jérémy Schiphof-Godart, Miranda Christianen, Henrike Westerveld, Lorne Luthart, Remi A. Nout, Mischa S. Hoogeman, Inger-Karine K. Kolkman-Deurloo

Erasmus MC Cancer Institute, University Medical Center, Department of Radiotherapy, Rotterdam, Netherlands

Purpose/Objective:

In intraoperative prostate HDR-BT treatment planning, transrectal ultrasound (TRUS) imaging is used for organ delineation and implant reconstruction. For correct dose delivery, the actual dwell positions need to precisely correspond to the treatment plan. However, TRUS-based implant reconstruction can be difficult and prone to uncertainties caused by needle shadowing and artifacts [1]. Electromagnetic tracking (EMT) has been proposed for detection of reconstruction errors [2]. The hybrid EMT/BT prototype Flexitron afterloading system with an EMT sensor integrated in the check cable (Elekta Brachytherapy, Veenendaal) has previously been validated for use in the clinical brachytherapy environment by phantom measurements [3], and was subsequently used in a prospective clinical study [4]. In the current study the EMT measurement reliability and added value in intraoperative TRUS based prostate HDR-BT is assessed. First, the EMT measurement stability was assessed using fixed reference sensor measurements. Second, needle reconstruction accuracy was investigated using EMT measurements. Next, reconstruction errors were identified and categorized by severity. Finally, the underlying causes of these errors were evaluated. Data of 14 patients with low-risk prostate cancer, treated with an HDR-BT fraction in the intraoperative setting, as part of a monotherapy treatment of 2x13.5Gy with a single implant, were evaluated. The implant consisted of plastic 6F ProGuide needles implanted via a perineal template under TRUS guidance. TRUS-based treatment planning was performed using Oncentra-Prostate. EMT measurements (Figure 1) were performed in an observational study (not used for clinical decision making). The study received approval from the institutional review board, and informed consent was obtained from all participants. In each measurement the EMT sensor followed a predefined track along the full length of each needle with dwell positions every 5 mm. In nine measurements an additional EMT reference sensor was placed in one of the catheters, measuring the relative displacement of the implant during the EMT measurements. The EMT measured dwell positions that were visible on the TRUS images were registered to the manually reconstructed dwell positions. The Euclidean distance (ED) between corresponding measured and reconstructed dwell positions was calculated. All needles with at least one dwell position having an ED >2 mm were regarded as erroneously reconstructed. Reconstruction errors were calculated per needle and graded following four levels of severity: Material/Methods:

• Minor: maximum ED ≥ 2 mm; median ED < 2 mm • Moderate: median ED ≥ 2 mm; maximum ED < 3 mm • Major: median ED ≥ 2 mm; maximum ED ≥ 3 mm; minimum ED < 3 mm • Severe: minimum ED ≥ 3 mm