In radiotherapy, precise dose delivery to the tumor while sparing healthy tissue is crucial for treatment success. This challenge is even greater in particle therapy, where sharp dose gradients increase sensitivity to anatomical uncertainties. Respiratory motion, common in thoracic and abdominal tumors, can cause significant displacements that compromise dose conformity and efficacy. Various motion mitigation techniques, such as gating, rescanning, and tracking, have been developed, with adaptive therapy emerging as a promising approach to adjust plans in response to anatomical changes. This is especially important in hypofractionated treatments, which use fewer, higher-dose sessions, amplifying motion effects per fraction. Accurate monitoring and verification of delivered dose in the presence of motion are essential enablers for precise and safe therapy. This work presents 5D-Real-time Ion DOse planning and delivery System (5D-RIDOS), a comprehensive system for real-time dose reconstruction in particle therapy, interfaced with a 4D-dose delivery system (4D-DDS). Unlike conventional approaches based on regular respiratory patterns and discrete 4D phases, 5DRIDOS integrates patient-specific motion modeling to continuously estimate time-resolved anatomy during treatment through a validated external–internal correlation model. Validation studies using virtual patient datasets and experimental evaluation demonstrated high anatomical accuracy and real-time processing capability. The reconstructed CT volumes showed strong spatial agreement with ground truth, with the target region reaching a maximum average boundary Hausdorff distance of 0.63 mm and a COM displacement of 1.06 mm. Dose reconstruction was completed within inter-spill times, with average gamma-index pass rates (3%/3mm) of 99% against detector measurements, including both regular and irregular motion scenarios. Moreover, 5D-RIDOS has been integrated within a comprehensive experimental framework developed to enable the delivery of 4D treatment plans using clinical-grade devices. This framework integrates real-time optical tracking system (OTS) with a 4D-DDS to demonstrate the feasibility of motion-compensated particle therapy in realistic clinical conditions. 5D-RIDOS serves as a continuous verification tool throughout treatment delivery, providing real-time dose reconstruction capabilities. Experimental validation with patient-optimized treatment plans demonstrated excellent dosimetric agreement between planned and detector-measured doses under regular motion conditions, particularly for regating and multi-phase 4D (MP4D) strategies (gamma-index pass rates of 99.6% and 97.3%, respectively, at 2%/2mm criteria). Under irregular motion scenarios, where discrepancies between planned and delivered doses naturally occurred, 5D-RIDOS continuously reconstructed the delivered dose in real-time, maintaining consistently high agreement with detector measurements (gamma-index pass rates always above 99% at 2%/2mm criteria) while simultaneously identifying deviations from the planned dose distribution. Finally, the framework’s adaptive capabilities were demonstrated through the first implementation of intrafractional plan adaptation, where dose information from one motion-degraded field was used to re-optimize a subsequent field within the same fraction. This proof-of-concept restored part of the intended dose distribution, improving gamma-index pass rates from 62.1% to 83.0% (3%/3mm). 5D-RIDOS and the complete framework represent a significant advancement toward clinically practical motion-aware particle therapy, providing the technological foundation for routine implementation of 4D treatments while ensuring treatment safety through comprehensive real-time dose verification

Real-time 4D-dose calculation to assess the efficacy of motion mitigation strategies

GALEONE, COSIMO
2025

Abstract

In radiotherapy, precise dose delivery to the tumor while sparing healthy tissue is crucial for treatment success. This challenge is even greater in particle therapy, where sharp dose gradients increase sensitivity to anatomical uncertainties. Respiratory motion, common in thoracic and abdominal tumors, can cause significant displacements that compromise dose conformity and efficacy. Various motion mitigation techniques, such as gating, rescanning, and tracking, have been developed, with adaptive therapy emerging as a promising approach to adjust plans in response to anatomical changes. This is especially important in hypofractionated treatments, which use fewer, higher-dose sessions, amplifying motion effects per fraction. Accurate monitoring and verification of delivered dose in the presence of motion are essential enablers for precise and safe therapy. This work presents 5D-Real-time Ion DOse planning and delivery System (5D-RIDOS), a comprehensive system for real-time dose reconstruction in particle therapy, interfaced with a 4D-dose delivery system (4D-DDS). Unlike conventional approaches based on regular respiratory patterns and discrete 4D phases, 5DRIDOS integrates patient-specific motion modeling to continuously estimate time-resolved anatomy during treatment through a validated external–internal correlation model. Validation studies using virtual patient datasets and experimental evaluation demonstrated high anatomical accuracy and real-time processing capability. The reconstructed CT volumes showed strong spatial agreement with ground truth, with the target region reaching a maximum average boundary Hausdorff distance of 0.63 mm and a COM displacement of 1.06 mm. Dose reconstruction was completed within inter-spill times, with average gamma-index pass rates (3%/3mm) of 99% against detector measurements, including both regular and irregular motion scenarios. Moreover, 5D-RIDOS has been integrated within a comprehensive experimental framework developed to enable the delivery of 4D treatment plans using clinical-grade devices. This framework integrates real-time optical tracking system (OTS) with a 4D-DDS to demonstrate the feasibility of motion-compensated particle therapy in realistic clinical conditions. 5D-RIDOS serves as a continuous verification tool throughout treatment delivery, providing real-time dose reconstruction capabilities. Experimental validation with patient-optimized treatment plans demonstrated excellent dosimetric agreement between planned and detector-measured doses under regular motion conditions, particularly for regating and multi-phase 4D (MP4D) strategies (gamma-index pass rates of 99.6% and 97.3%, respectively, at 2%/2mm criteria). Under irregular motion scenarios, where discrepancies between planned and delivered doses naturally occurred, 5D-RIDOS continuously reconstructed the delivered dose in real-time, maintaining consistently high agreement with detector measurements (gamma-index pass rates always above 99% at 2%/2mm criteria) while simultaneously identifying deviations from the planned dose distribution. Finally, the framework’s adaptive capabilities were demonstrated through the first implementation of intrafractional plan adaptation, where dose information from one motion-degraded field was used to re-optimize a subsequent field within the same fraction. This proof-of-concept restored part of the intended dose distribution, improving gamma-index pass rates from 62.1% to 83.0% (3%/3mm). 5D-RIDOS and the complete framework represent a significant advancement toward clinically practical motion-aware particle therapy, providing the technological foundation for routine implementation of 4D treatments while ensuring treatment safety through comprehensive real-time dose verification
26-nov-2025
Inglese
GRAEFF, CHRISTIAN
VIGNATI, Anna
Università degli Studi di Torino
File in questo prodotto:
File Dimensione Formato  
Tesi_Cosimo_Galeone.pdf

accesso aperto

Licenza: Tutti i diritti riservati
Dimensione 15.8 MB
Formato Adobe PDF
15.8 MB Adobe PDF Visualizza/Apri

I documenti in UNITESI sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.

Utilizza questo identificativo per citare o creare un link a questo documento: https://hdl.handle.net/20.500.14242/312678
Il codice NBN di questa tesi è URN:NBN:IT:UNITO-312678