Dosimetric Impact of the CT-ED Curve in Radiotherapy of Pediatric Patients

Mikić, Milica (author), Simončič, Urban (mentor), Peterlin, Primož (comentor)

The electron density on the tomogram is normalized in CT-numbers. An experimentally obtained CT-ED curve is used to determine the electron density (ED). The relationship between the CT number and the electron density of the material/tissue depends on the voltage of the cathode tube. In radiotherapy, we usually use the relationship between the CT-number and the electron density of the substance obtained at 120 kV, which corresponds to most cases. Pediatric patients are an important exception. They are often imaged at lower tube voltages (70 kV or 80 kV), which contributes to better soft tissue contrast but also results in a somewhat lower received dose. Nevertheless, a standard conversion table between CT-number and electron density is used in treatment planning, resulting in an incorrect (too high) electron density being assigned to the CT-number for materials or tissues with higher electron density. The dosimetric impact of this error was calculated in this study. Pediatric patients under 14 years of age treated at the Department of External Radiation Therapy of the Oncology Institute in Ljubljana were included in this study. The group size was 32 patients. The cases of pediatric patients imaged with Siemens Somatom Definition AS and Philips Brilliance Big Bore scanners were collected, and their CT images were exported and anonymized. The CT-ED tables for different tube voltages were obtained from measurements on phantoms. A computer program was written to read images in DICOM format and correct them with respect to different X-ray tube voltages (80, 90, 100, and 120 kV). The modified CT images were imported into the radiation planning system, the existing clinical radiation plan was mapped onto them, and the spatial distribution of absorbed dose was calculated. The absorbed dose spatial distributions for each patient (actually used and calculated to 120 kV) were exported and anonymized. Finally, differences in absorbed dose calculated on the original and modified CT images were analysed. For low densities up to 1.1 g/cm3, Hounsfield unit (HU) differences of 0–3% were found, but for densities in the range above 1.1 g/cm3, HU variations as a function of X-ray tube voltage of up to 31% were found. The maximum calculated difference in absorbed dose in the tumour (i.e. planning target volume) proved to be less than 1%, while in critical organs it was less than 2.2%. Therefore, the error due to the use of an incorrect CT-ED conversion curve does not critically affect the dose distribution.

• Language: English
• Keywords: radiotherapy of pediatric patients, computed tomography, CT-ED curve, Hounsfield units, dosimetry, treatment planning system, dose-volume histogram, gamma index analysis
• Work type: Master's thesis/paper
• Typology: 2.09 - Master's Thesis
• Organization: FMF - Faculty of Mathematics and Physics
• Year: 2023
• PID: 20.500.12556/RUL-143859
• COBISS.SI-ID: 138119683
• Publication date in RUL: 15.01.2023

Secondary language

• Language: Slovenian
• Title: Dozimetrični vpliv CT-ED krivulje pri radioterapiji pediatričnih bolnikov

Abstract

Elektronska gostota na tomogramu je normalizirana v Hounsfieldovih enotah, oz. CT-številih. Za določitev elektronske gostote se uporablja eksperimentalno dobljena CT-ED krivulja. Zveza med CT-številom in elektronsko gostoto snovi je odvisna od pospeševalne napetosti. V radioterapiji večinoma uporabljamo zvezo med CTštevilom in elektronsko gostoto snovi, dobljeno pri 120 kV, kar bolj ali manj ustreza večini primerov. Pomembna izjema so pediatrični bolniki. Pogosto jih slikamo pri nižjih pospeševalnih napetostih (70 kV ali 80 kV), kar po eni strani pripomore k boljši ločljivosti anatomskih struktur, po drugi pa k nekoliko manjši prejeti slikovni absorbirani dozi. Kljub temu se pri načrtovanju zdravljenja uporablja standardna pretvorbena tabela med CT-številom in elektronsko gostoto, zaradi česar je za materiale ali tkiva z večjo elektronsko gostoto CT-številu dodeljena nepravilna (previsoka) elektronska gostota. Dozimetrični vpliv te napake je bil izračunan v tej študiji. V raziskavo so bili vključeni pediatrični bolniki, mlajši od 14 let, zdravljeni na Oddelku za teleradioterapijo Onkološkega inštituta v Ljubljani. V skupini je bilo 32 bolnikov. Zbrani so bili primeri pediatričnih bolnikov, posnetih s skenerji Siemens Somatom Definition AS in Philips Brilliance Big Bore, njihove CT slike pa so bile izvožene in anonimizirane. Tabele CT-ED za različne pospeševalne napetosti so bile pridobljene z meritvami na fantomih. Napisan je bil računalniški program za branje slik v formatu DICOM in njihovo korekcijo glede na različne pospeševalne napetosti (80, 90, 100 in 120 kV). Spremenjene CT slike smo uvozili v sistem za načrtovanje obsevanj, vanje preslikali obstoječi načrt kliničnega obsevanja in izračunali prostorsko porazdelitev absorbirane doze. Prostorske porazdelitve absorbirane doze za vsakega bolnika (dejansko uporabljene in izračunane na 120 kV) so bile izvožene in anonimizirane. Na koncu so bile analizirane razlike v absorbirani dozi, izračunani na originalni in spremenjeni CT sliki. Za nizke gostote do 1.1 g/cm3 so bile ugotovljene razlike v Hounsfieldovih enotah (HU) od 0–3%, toda za gostote v območju nad 1.1 g/cm3 so bile HU razlike, kot funkcija pospeševalne napetosti rentgenske cevi, do 31%. Največja izračunana razlika v absorbirani dozi v tumorju (tj. planirnem tarčnem volumnu) je bila manjša od 1 %, v kritičnih organih pa manj kot 2.2 %. Od tod sledi, da napaka zaradi uporabe neustrezne CT-ED pretvorbene krivulje ne vpliva kritično na dozno porazdelitev.

• Keywords: radioterapija pediatričnih bolnikov, računalniška tomografija, CT-ED krivulja, Hounsfield enote, dozimetrija, sistem za načrtovanje zdravljenja, dozno-volumski histogram, analiza indeksa gama