Background: Computed tomography (CT) imaging is a diagnostic imaging technique involving significant ionizing exposure which dose estimation to ensure patient safety. Traditional methods using conversion coefficients are widely used but may lack precision. Monte Carlo simulations offer a more patient-specific alternative for dose estimation. Objective: This study aims to compare the accuracy and reliability of traditional dose coefficient methods versus Monte Carlo simulations in estimating the effective radiation dose estimation in CT scans. Method: We analyzed data from 56 patients (45M; 11F; mean age: 62.5+/-15.04) referred to intensive care unit and who underwent multiple CT scans. Effective dose was estimated by using conversion coefficients and dose-length-product and Monte Carlo simulations. Agreement between the two methods was assessed through percentage analysis, Bland-Altman plots, and Pearson correlation. Those effective dose values comprised within 15% of Monte Carlo values were considered concordant. Result: Monte Carlo simulations provided more accurate and patient-specific dose estimates than coefficient method. The coefficient method underestimated radiation doses, particularly at higher levels, with only 42% (24/56) of estimates falling within the acceptable ±15% range. The Bland-Altman plot revealed a systematic bias in traditional methods indicating a consistent underestimation of effective doses by the coefficient method compared to the Monte Carlo simulations. This bias becomes more pronounced at higher dose levels. While Pearson correlation showed a strong linear relationship (r = 0.774) between the two methods.
Introduzione: La tomografia computerizzata (TC) è una tecnica di imaging diagnostico; tuttavia, comporta un'esposizione significativa a radiazioni ionizzanti, rendendo necessaria una stima accurata della dose per garantire la sicurezza del paziente. I metodi tradizionali basati su coefficienti di conversione sono comunemente utilizzati, ma possono essere meno precisi. Le simulazioni Monte Carlo rappresentano un'alternativa più specifica e precisa per la stima della dose efficace. Obiettivo: Lo scopo di questo studio è confrontare l'accuratezza e l'affidabilità dei metodi tradizionali basati sui fattori dei coefficienti con le simulazioni Monte Carlo per la stima della dose efficace nelle scansioni TC. Metodo: Sono stati analizzati i dati di 56 pazienti sottoposti a diverse scansioni TC. Le dosi efficaci sono state calcolate sia con il metodo tradizionale che con le simulazioni Monte Carlo. La precisione è stata valutata utilizzando analisi percentuali, diagramma di Bland-Altman e il coefficiente di correlazione di Pearson. Sono stati considerati accettabili i casi in cui il metodo tradizionale rientrava entro un intervallo di ±15% rispetto ai valori Monte Carlo, ed i dati sono stati analizzati per identificare tendenza e bias. Risultati: Le simulazioni Monte Carlo hanno fornito stime della dose più accurate e personalizzate. Il metodo dei coefficienti, invece, ha sottostimato le dosi di radiazione, in particolare a livelli più elevati, con solo il 42% (24/56) delle stime rientranti nell’intervallo accettabile di ±15%. Il diagramma di Bland-Altman ha rivelato un bias sistematico nei metodi tradizionali, indicando una sottostima costante delle dosi efficaci da parte del metodo dei coefficienti rispetto alle simulazioni Monte Carlo. Questo bias diventa più evidente a livelli di dose più elevati. mentre la correlazione di Pearson ha mostrato una forte relazione lineare (r = 0,774) tra i due metodi.
Effective Dose Calculation in Critical Patients: A Comparative Analysis between Monte Carlo simulations and coefficient calculations
KANSO, ISSAM
2023/2024
Abstract
Background: Computed tomography (CT) imaging is a diagnostic imaging technique involving significant ionizing exposure which dose estimation to ensure patient safety. Traditional methods using conversion coefficients are widely used but may lack precision. Monte Carlo simulations offer a more patient-specific alternative for dose estimation. Objective: This study aims to compare the accuracy and reliability of traditional dose coefficient methods versus Monte Carlo simulations in estimating the effective radiation dose estimation in CT scans. Method: We analyzed data from 56 patients (45M; 11F; mean age: 62.5+/-15.04) referred to intensive care unit and who underwent multiple CT scans. Effective dose was estimated by using conversion coefficients and dose-length-product and Monte Carlo simulations. Agreement between the two methods was assessed through percentage analysis, Bland-Altman plots, and Pearson correlation. Those effective dose values comprised within 15% of Monte Carlo values were considered concordant. Result: Monte Carlo simulations provided more accurate and patient-specific dose estimates than coefficient method. The coefficient method underestimated radiation doses, particularly at higher levels, with only 42% (24/56) of estimates falling within the acceptable ±15% range. The Bland-Altman plot revealed a systematic bias in traditional methods indicating a consistent underestimation of effective doses by the coefficient method compared to the Monte Carlo simulations. This bias becomes more pronounced at higher dose levels. While Pearson correlation showed a strong linear relationship (r = 0.774) between the two methods.File | Dimensione | Formato | |
---|---|---|---|
TESI DEF.pdf
accesso riservato
Dimensione
1.61 MB
Formato
Adobe PDF
|
1.61 MB | Adobe PDF |
The text of this website © Università degli studi di Padova. Full Text are published under a non-exclusive license. Metadata are under a CC0 License
https://hdl.handle.net/20.500.12608/80412