High dose rate brachytherapy treatment verification using a flat panel detector

Smith, R 2017, High dose rate brachytherapy treatment verification using a flat panel detector, Doctor of Philosophy (PhD), Science, RMIT University.

Document type: Thesis
Collection: Theses

Attached Files
Name Description MIMEType Size
Smith.pdf Thesis application/pdf 6.64MB
Title High dose rate brachytherapy treatment verification using a flat panel detector
Author(s) Smith, R
Year 2017
Abstract High dose rate (HDR) brachytherapy treatments are usually delivered in large dose fractions and have the clinical advantage of highly conformal dose distributions due to the steep dose gradient produced by the 192Ir source. The routine use of 3D imaging for treatment planning enables clinical teams to finely optimise the dose distribution around the defined target while limiting dose to the surrounding organs at risk. A significant challenge in brachytherapy is to ensure the dose is delivered to the patient as planned, which can be challenging due to factors that impact the accuracy of dose delivery. Due to the high degree of manual processes in brachytherapy, the relative risk of treatment delivery error is high when compared to other radiotherapy modalities. Additionally, interstitial and intracavitary brachytherapy suffer from anatomical motion and swelling due to catheter (or applicator) implant trauma.

There are two fundamental ways to verify a HDR brachytherapy treatment delivery: (i) verify the source dwell positions and times are as per the treatment plan, or (ii) perform a measurement in vivo with a dosimeter. Although reported in many small patient studies, in vivo dosimetry (IVD) has many limitations (e.g. detector position uncertainty and limited sampling) making the interpretation of results for treatment verification difficult. These challenges may be the reason for the limited routine application of IVD as a treatment verification technique. Since the treatment plan is a planned set of dwell positions and times, the former approach has the potential to verify the dose over the entire treatment volume.

This thesis addresses the need to improve the methodology for treatment verification in HDR brachytherapy. This work aims to establish a verification technique that can be used routinely in the clinical environment, not impact the patient and provide data that can be confidently interpreted to verify the entire treatment delivery. To achieve this, a novel approach to treatment verification was investigated, avoiding the challenges of directly measuring dose. Measurements of the source position, during the treatment delivery (source tracking) were made, enabling direct comparison with the treatment plan for verification. Additionally, a method to establish a structured approach for performing this treatment verification process was accomplished, with the objective to enable routine use and widespread uptake of this process. The overall goal of this novel verification approach was to improve the quality of treatment delivery and patient safety in HDR brachytherapy.

To investigate this new approach, a flat panel detector (FPD) was employed as the measurement device. The detector, originally designed for use as an electronic portal imaging device, was characterised for use with an 192Ir brachytherapy source. The FPD response and the image acquisition timing were investigated to demonstrate its capability for this work. The images of the response to the 192Ir source, acquired with the FPD, were interpreted by a range of algorithms, extracting metrics that could be correlated with the position of the source. A concept for integration into a clinical environment was developed, by placing the FPD in the treatment couch, immediately below the target volume. The potential for the brachytherapy implant to displace due to anatomical and other influences was addressed by performing pre-treatment image verification. An imaging geometry was established, allowing registration with the treatment plan, enabling identification and quantification of implant displacement (in the treatment bunker) immediately prior to treatment delivery. The relationship established between the measurement frame of reference and the treatment plan permitted direct comparison of the measured dwell positions with the planned dwell positions for verification of treatment delivery. Treatment delivery metrics were developed to detect the occurrence of a treatment error, and based on the unique signature of the error, identification of the error source was possible. This concept of treatment verification was transferred into the clinical environment and a patient measurement was performed to understand the challenges of clinical implementation.

The FPD responded to the 192Ir source, for a range of clinically relevant distances (20 to 200 mm) away from the FPD despite the low dose rates and the changing photon spectrum. The image acquisition time was one image capture every 1.8 seconds, and although not designed for this application, the FPD was adequate to perform this proof of principle work. Using a range of algorithms, the images acquired by the FPD were processed to determine the source position. It was determined that a centre of mass approach was the most accurate method (x and y s.d. 0.3 and 0.1 mm, up to 200 mm from the FPD imaging plane) to determine source position in the 2D plane of the FPD. The influence of inhomogeneities and finite phantom geometry were quantified relative to their influence on the accuracy of determining the source position when applied in a clinical scenario. A structured approach to pre-treatment imaging was developed, with a robust method to perform a 3D reconstruction of the implant in the treatment bunker using a ‘shift image’ technique. A registration between the treatment planning system (TPS) and the measurement space (FPD) was established allowing quantitative evaluation of the implant changes since treatment planning imaging. Pre-treatment imaging was capable of identifying catheter displacements in the order of 2.0 mm with a confidence of 95%. Identification of a treatment delivery error was possible with the use of metrics that when combined define an error ‘signature’ that suggest the source of the error. The absolute relationship between the measurement space and the TPS allow error trapping to identify errors that would otherwise go undetected, for example an incorrect channel length definition error. This verification approach was applied successfully in a clinical setting. Pre-treatment imaging allowed confirmation that the implanted catheters had not significantly displaced prior to treatment and source tracking results confirmed the treatment was delivered as planned. The clinical implementation had minimal impact on the workflow, increasing the patient setup (and imaging) time by only 15 minutes while not adding any additional time to the radiation dose delivery portion of the treatment.

This initial work using a FPD for treatment verification in HDR brachytherapy has highlighted the benefits of this approach. This novel approach provides multiple layers of verification, including pre-treatment imaging (in 2D and 3D) to identify potential sources of error prior to treatment delivery. Source tracking, in conjunction with pre-treatment imaging, provides quantitative verification of the entire treatment delivery, currently not possible with other methods. This approach establishes a new standard of verification which has the potential to improve the quality of treatment delivery and improves patient safety in HDR brachytherapy.
Degree Doctor of Philosophy (PhD)
Institution RMIT University
School, Department or Centre Science
Subjects Medical Physics
Keyword(s) HDR Brachytherapy
Treatment Verification
Pre-Treatment Imaging
Catheter Displacement
Version Filter Type
Access Statistics: 122 Abstract Views, 101 File Downloads  -  Detailed Statistics
Created: Tue, 24 Apr 2018, 13:20:14 EST by Keely Chapman
© 2014 RMIT Research Repository • Powered by Fez SoftwareContact us