Brachytherapy treatment verification using gamma radiation from the internal treatment source combined with an imaging panel-a phantom study.
Gabriel Paiva FonsecaTeun Pieter van WagenbergR VonckenM PodestaC van BeverenE van LimbergenL LutgensB VannesteM BerbeeB ReniersF VerhaegenPublished in: Physics in medicine and biology (2021)
Brachytherapy has an excellent clinical outcome for different treatment sites. However,in vivotreatment verification is not performed in the majority of hospitals due to the lack of proper monitoring systems. This study investigates the use of an imaging panel (IP) and the photons emitted by a high dose rate (HDR)192Ir source to track source motion and obtain some information related to the patient anatomy. The feasibility of this approach was studied by monitoring the treatment delivery to a 3D printed phantom that mimicks a prostate patient. A 3D printed phantom was designed with a template for needle insertion, a cavity ('rectum') to insert an ultrasound probe, and lateral cavities used to place tissue-equivalent materials. CT images were acquired to create HDR192Ir treatment plans with a range of dwell times, interdwell distances and needle arrangements. Treatment delivery was verified with an IP placed at several positions around the phantom using radiopaque markers on the outer surface to register acquired IP images with the planning CT. All dwell positions were identified using acquisition times ≤0.11 s (frame rates ≥ 9 fps). Interdwell distances and dwell positions (in relation to the IP) were verified with accuracy better than 0.1 cm. Radiopaque markers were visible in the acquired images and could be used for registration with CT images. Uncertainties for image registration (IP and planning CT) between 0.1 and 0.4 cm. The IP is sensitive to tissue-mimicking insert composition and showed phantom boundaries that could be used to improve treatment verification. The IP provided sufficient time and spatial resolution for real-time source tracking and allows for the registration of the planning CT and IP images. The results obtained in this study indicate that several treatment errors could be detected including swapped catheters, incorrect dwell times and dwell positions.
Keyphrases
- high dose
- image quality
- prostate cancer
- computed tomography
- deep learning
- dual energy
- magnetic resonance
- combination therapy
- low dose
- machine learning
- radiation therapy
- optical coherence tomography
- convolutional neural network
- high resolution
- positron emission tomography
- ultrasound guided
- radiation induced
- liquid chromatography
- social media