Usporedba algoritama za izračun raspodjele doze u radioterapiji i nova metoda verifikacije Monte Carlo algoritama u nehomogenostima

doctoral thesis

• Title (english): Comparison of dose calculation algorithms for radiotherapy and novel method to evaluate Monte Carlo based photon dose calculation algorithms in inhomogeneities
• Author: Đeni Smilović Radojčić
• Mentor: Slaven Jurković (mentor)
• Committee member: Marija Majer (predsjednik povjerenstva)
• Committee member: Slaven Jurković (član povjerenstva)
• Committee member: Željka Knežević Medija (član povjerenstva)
• Granter: University of Zagreb; Faculty of Science; (Department of Physics); Zagreb
• Defense date and country: 2018-09-28, Croatia
• Scientific / art field, discipline and subdiscipline: NATURAL SCIENCES; Physics
• Universal decimal classification (UDC): 53 - Physics

Abstract

U doktorskom radu su prikazani metodologija i rezultati dozimetrijskog vrednovanja sustava za izračun i optimizaciju raspodjele apsorbirane doze Elekta Monaco, koji za izračunavanje apsorbirane doze koristi algoritam koji se temelji na Monte Carlo simulaciji. Sustav pruža mogućnost izračuna apsorbirane doze na dva načina, kao doza u sredstvu i kao doza u vodi. Dozimetrijsko vrednovanje sustava za izračun i optimizaciju raspodjele apsorbirane doze provedeno je u fantomima različitih vrsta, tako da je izmjerena raspodjela apsorbirane doze uspoređivana s izračunatom. Za mjerenje apsorbirane doze korištene su ionizacijske komore različitih efektivnih mjernih volumena i dvodimenzionalni detektor sastavljen od matrice ionizacijskih komora. U sredstvima gustoće veće od gustoće vode uočene su statistički značajne razlike između dvije metode izračuna. Zbog uočenih statistički značajnih razlika u svrhu dodatne analize je osmišljeno više mjernih/računskih geometrija. Na temelju analize rezultata predložena je metoda vrednovanja sustava za izračun i optimizaciju raspodjele apsorbirane doze u sredstvima gustoće veće od gustoće vode. Kao nadopuna vrednovanju računalnog algoritma u točki provedeno je dvodimenzionalno dozimetrijsko vrednovanje raspodjele apsorbirane doze. Osmišljene su različite nehomogene mjerne i računske geometrije kako bi snop X-zraka na putu do ionizacijskih komora prošao kroz sredstva različitih gustoća. Isti trend rezultata koji su dobiveni u semi-antropomorfnom fantomu dobiveni su i analizom razlike izračunate apsorbirane doze koristeći dvije metode izračuna u stvarnim kliničkim situacijama. Naime, iako odstupanja između dvije metode izračuna u mekom tkivu nisu bila statistički značajna, u organima rizika izgrađenih od koštanih struktura (mandibula, kohlea) razlika između dvije metode izračuna je statistički značajna.

Abstract (english)

Purpose: This doctoral dissertation presents the methodology and results for the dosimetric evaluation of absorbed dose distribution calculation and optimization system Monaco, which uses a Monte Carlo simulation based algorithm. Monte Carlo based treatment planning system (TPS) are the most accurate approach for particle transport. They can calculate absorbed dose in terms of dose to media, Dm, as well as in terms of dose to water, Dw, which is retrospectively converted from Dm. The purpose of our study was to analyse differences between dose-tomedium and dose-to-water calculation options for Monaco TPS (Elekta, Sweden) and to propose a supplement to the existing verification methodology to make the verification of Monte Carlo based systems used for radiation oncology unambiguous and straightforward. Methods: Dosimetric evaluation for the two calculation modes was performed using several homogeneous phantoms, a semi-anthropomorphic and an anthropomorphic phantom. Measured absorbed dose distributions were compared to calculated absorbed dose distributions. Ionization chambers with different effective volumes and a two-dimensional detector with an ionization chamber array were used for measuring the absorbed dose. Additionally, measured absorbed dose distributions were compared to absorbed dose distributions calculated with two analytic calculation algorithms. As a supplement to the calculation algorithm evaluation in a point, a two dimensional dosimetric evaluation of absorbed dose distribution was performed. Different inhomogeneous measuring and computing geometries were designed to allow X-ray beams to pass through media of different densities on their path to the array of small volume ionization chambers. Measured absorbed dose distributions were compared to those calculated using gamma method which assesses the difference in dose for low dose gradient regions and spatial component for high dose gradient regions. Furthermore, we investigated the effect of dose-to-water and dose-to-medium calculation options built in Elekta Monaco TPS on intensity modulated radiotherapy (IMRT) dose distributions in clinical situations. Results: For a water equivalent medium and low density medium differences were statistically insignificant and it was confirmed that the most accurate absorbed dose distributions were calculated using an Monte Carlo based calculation algorithm. The statistically significant differences were observed for high density media between measured and calculated absorbed dose values and therefore several measuring/calculation geometries were designed for additional analysis. To overcome potential ambiguity on the correctness of the verification results in high-density media, we proposed a supplement to current comprehensive verification methodology for performance of algorithm built in the Monte Carlo based treatment planning systems. It is based on the physical principles which are ground for the concept of dose measurements of high energy X-ray beams used for radiotherapy. With this addendum Monte Carlo based calculation algorithms can be verified in described geometry for high-density media with higher level of confidence. Considerable difference among dose distributions calculated using respective calculation option obtained in semi-anthropomorphic phantom was also reported in clinical cases. Even though the discrepancies in tissue equivalent media (all target volumes and most of organs at risk) were not statistically significant, in bony structured organs at risk (mandible, cochlea) the discrepancies between two calculation modes was shown to be statistically significant. Conclusions: This dissertation presents the results of the dosimetric evaluation of the calculation algorithm which uses Monte Carlo simulation for absorbed dose calculation, built in a system for calculation and optimization of absorbed dose distribution Elekta Monaco. A supplement to current comprehensive verification methodology for performance of algorithm built in the Monte Carlo based treatment planning systems was proposed. Although proposed addendum to the verification methodology is not the answer to the dilemma which calculation option shall be used for radiotherapy treatment prescription, it provides additional tools for the verification of the TPS calculation algorithms based on the physics behind.

• Keywords: apsorbirana doza; dozimetrija; Monte Carlo simulacija; doza u vodi; doza u sredstvu
• Keywords (english): absorbed dose; dosimetry; Monte Carlo simulation; dose to water; dose to media
• Language: croatian
• URN:NBN: urn:nbn:hr:217:240736
• Study programme: Title: Physics; Study programme type: university; Study level: postgraduate; Academic / professional title: doktor/doktorica znanosti, područje prirodnih znanosti, polje fizika (doktor/doktorica znanosti, područje prirodnih znanosti, polje fizika)
• Type of resource: Text
• Extent: 102 str.
• File origin: Born digital
• Access conditions: Open access
• Created on: 2018-10-10 11:51:56