Optimization algorithm of a high dose-rate brachytherapy treatment plan for cervical cancer patients in Zambia.
Loading...
Date
2023
Authors
Kaniini, Mannah
Journal Title
Journal ISSN
Volume Title
Publisher
The University of Zambia
Abstract
The increase in availability of diagnostic tools and increase in awareness of cancer disease, has enabled the diagnosis of new cancer cases that were not previouslycaptured in Zambia. Today, the Cancer Diseases Hospital (CDH) treats over 3,200 cancer patients annually, a number significantly
higher than the initial projection of 1,200 when CDH was established. Approximately 60% of these patients require Radiotherapy (RT), and Brachytherapy (BT) is an essential component for treating cervical cancer. At CDH, BT treatment planning employs a computerized treatment planning
system based on the report by Task Group 43 (TG-43) of American Association of Physicists in Medicine (AAPM) that generates radiation dose distributions for optimal treatment delivery. Although individualized plans can be developed from standard treatment plans, high patient volumes necessitate generating multiple standard plans tailored to suit most patient profiles. However, there is a need to verify the accuracy of these standard plans. This study aims at refining strategic decision points for passing or failing the standard treatment plan used at CDH by developing a mathematical modeling algorithm that evaluates current practices in optimizing. This is achieved by developing brachytherapy standard treatment plans. Using a retrospective study design spanning three years (2019-2021), we reviewed 254 treatment plans and collected data by reviewing source details from radioactive source parameters and dose-volume histogram markers to calculate various dose metrics necessary for quality assurance purposes. The study used some dose metrics to develop an algorithm capable of determining whether a given plan passes or fails
based on quantitative evaluation criteria. While values for the Conformity Index (COIN), Conformity Number (CN), and Coverage Index (CI) were better in 2019 compared with other years; Dose Non-Uniformity Ratio (DNR) and Dose Homogeneity Index (DHI) remained low throughout all years reviewed. Nonetheless, most analyzed treatments had acceptable values for COIN, CI, and CN with mean values ranging between 0.86 (±0.09) to 0 .93 (±0 .05). Ideal case entails that these indices are of the value 1 or close to 1 except for DNR where a value of 0 is ideal. Unfortunately, none of the CDH standard plans examined passed the set criteria (i.e. 5 indices passing out of the 5) because none of them met standards for both DNR and DHI indices whose average values ranged between 0.34 to 0.43 for DHI and 0.57 to 0.66 for DNR. Therefore, achieving treatment plans that meet optimum balance among these indices require that organs at risk are not exposed beyond their tolerance limit but, providing maximum coverage of clinical target volume should be considered. This Algorithm will serve as an institutional protocol devised for acceptability criteria of a treatment plan at CDH.
Description
Thesis of Master of Science in Physics.