Proton therapy is a highly advanced form of external beam radiotherapy, which uses proton beams to treat cancers. Protons are particularly suited for radiotherapy because heavy particles, interacting with matter, deposit most of their energy at the end of their track. It allows for well-conformed dose distributions to the target area while minimizing exposure to surrounding healthy tissues and organs at risk. In addition, ongoing research reveals that protons have distinct biological, immunological, and clinical effects compared to photons, beyond just dose distribution differences. X-ray CT images are currently the standard for proton beam planning, but they have limitations due to the need to convert electron density maps into stopping power ratios (SPR) maps. CT numbers (HU values) noise and the lack of a consistent conversion to SPR lead to imperfect dose calculations and introduce density uncertainties. Additionally, CT scans only capture a single snapshot of patient anatomy, while daily variations during treatment cause setup uncertainties. Treatment planning systems (TPSs) are designed for account density and setup uncertainties through robust optimization. Density and setup uncertainty values are not uniquely defined. They are institution-specific, depending on the utilized treatment modalities, the clinical experience and the target anatomic localization. Density uncertainty typically ranges within 3% and 3.5%, whereas, setup uncertainty varies from 1 mm up to 5mm. Within the scientific community, there is a strong interest in reducing the density uncertainties. The advent of Dual-Energy CT (DECT) scanners has introduced new options to increase the SPR conversion accuracy compared to Single-Energy CT (SECT). The Siemens Syngo.via software is an available commercial solution to perform the DE raw image processing, generating Dual Energy Monoenergetic Plus images. In addition, the software includes a tool, DirectSPR, able to directly calculate voxel by voxel the proton SPR, without resorting to HU conversion. This work aims to evaluate the impact of HU to SPR conversion methods on TPS proton range estimation. HU to SPR conversion relationships were derived from CT calibrations. The following methods were implemented and evaluated: the SECT tissue-substitute method and, on the other hand, the SECT and Dual Energy Monoenergetic Plus Stoichiometric calibration methods. In different scenarios, the proton range was directly measured with a multi-layer ion chamber and compared with the one derived from TPS dose calculation on DirectSPR maps and SECT and DECT images. Firstly, tissue-surrogate solid plugs were individually irradiated with a single pencil beam of 226 MeV proton. DirectSPR-derived proton range showed a better agreement with measurements than CT calibration-based methods differing at most by 1% (SECT and 75KeV Monoenergetic Plus with the stoichiometric calibration respectively by 6.5% and 4.5%), for all the plugs except for high-density ones. All the TPS calculated proton ranges significantly differed (up to 32%) from the measurement for the He Cortical Bone plug. This result will require additional investigation. Plug results can not be fully translated to human tissues, since they have different chemical compositions. For this purpose, animal organ samples were used. They were irradiated with single pencil beams of different energies (100 MeV, 150 MeV and 226 MeV) and with a simple intensity modulated proton therapy (IMPT) plan (a dose cube). Three irradiation points were identified on each sample surface and the proton range was probed at each of them. Proton range differences between measures and TPS-derived ones had no dependency on beam energy. However, the results of this investigation do not confirm the high accuracy in SPR assignment of DirectSPR than SECT and DECT calibration-based methods, indeed, proton range differences with measures were very close within the various approaches. The density uncertainty was evaluated on the data derived from IMPT irradiation on animal organ samples, given the closeness to clinical treatment conditions. A value of 2% could be taken as density uncertainty associated with the use of SECT and Monoenergetic Plus75keV image with the relative stoichiometric calibration curve in proton dose calculation, while the value obtained for the DirectSPR was higher, 2.8%. The high standard deviations of the multiple irradiation point measurements suggest the presence of residual setup errors affecting the results and the necessity to improve setup repeatability for further investigations. The results of this study confirm the feasibility of the use of DirectSPR and Monoenergetic Plus images generated by the Syngo.via software as input to the Raystation TPS for proton dose calculation. Although preliminary, the results of this study support the reduction of the 3.5% clinical density uncertainty enabled by the implementation of dual-energy CT as suggested by many authors.
Evaluation of the proton range accuracy using Single and Dual Energy Computed Tomography / D. Alio. - (2024 Nov 14).
Evaluation of the proton range accuracy using Single and Dual Energy Computed Tomography
D. Alio
2024
Abstract
Proton therapy is a highly advanced form of external beam radiotherapy, which uses proton beams to treat cancers. Protons are particularly suited for radiotherapy because heavy particles, interacting with matter, deposit most of their energy at the end of their track. It allows for well-conformed dose distributions to the target area while minimizing exposure to surrounding healthy tissues and organs at risk. In addition, ongoing research reveals that protons have distinct biological, immunological, and clinical effects compared to photons, beyond just dose distribution differences. X-ray CT images are currently the standard for proton beam planning, but they have limitations due to the need to convert electron density maps into stopping power ratios (SPR) maps. CT numbers (HU values) noise and the lack of a consistent conversion to SPR lead to imperfect dose calculations and introduce density uncertainties. Additionally, CT scans only capture a single snapshot of patient anatomy, while daily variations during treatment cause setup uncertainties. Treatment planning systems (TPSs) are designed for account density and setup uncertainties through robust optimization. Density and setup uncertainty values are not uniquely defined. They are institution-specific, depending on the utilized treatment modalities, the clinical experience and the target anatomic localization. Density uncertainty typically ranges within 3% and 3.5%, whereas, setup uncertainty varies from 1 mm up to 5mm. Within the scientific community, there is a strong interest in reducing the density uncertainties. The advent of Dual-Energy CT (DECT) scanners has introduced new options to increase the SPR conversion accuracy compared to Single-Energy CT (SECT). The Siemens Syngo.via software is an available commercial solution to perform the DE raw image processing, generating Dual Energy Monoenergetic Plus images. In addition, the software includes a tool, DirectSPR, able to directly calculate voxel by voxel the proton SPR, without resorting to HU conversion. This work aims to evaluate the impact of HU to SPR conversion methods on TPS proton range estimation. HU to SPR conversion relationships were derived from CT calibrations. The following methods were implemented and evaluated: the SECT tissue-substitute method and, on the other hand, the SECT and Dual Energy Monoenergetic Plus Stoichiometric calibration methods. In different scenarios, the proton range was directly measured with a multi-layer ion chamber and compared with the one derived from TPS dose calculation on DirectSPR maps and SECT and DECT images. Firstly, tissue-surrogate solid plugs were individually irradiated with a single pencil beam of 226 MeV proton. DirectSPR-derived proton range showed a better agreement with measurements than CT calibration-based methods differing at most by 1% (SECT and 75KeV Monoenergetic Plus with the stoichiometric calibration respectively by 6.5% and 4.5%), for all the plugs except for high-density ones. All the TPS calculated proton ranges significantly differed (up to 32%) from the measurement for the He Cortical Bone plug. This result will require additional investigation. Plug results can not be fully translated to human tissues, since they have different chemical compositions. For this purpose, animal organ samples were used. They were irradiated with single pencil beams of different energies (100 MeV, 150 MeV and 226 MeV) and with a simple intensity modulated proton therapy (IMPT) plan (a dose cube). Three irradiation points were identified on each sample surface and the proton range was probed at each of them. Proton range differences between measures and TPS-derived ones had no dependency on beam energy. However, the results of this investigation do not confirm the high accuracy in SPR assignment of DirectSPR than SECT and DECT calibration-based methods, indeed, proton range differences with measures were very close within the various approaches. The density uncertainty was evaluated on the data derived from IMPT irradiation on animal organ samples, given the closeness to clinical treatment conditions. A value of 2% could be taken as density uncertainty associated with the use of SECT and Monoenergetic Plus75keV image with the relative stoichiometric calibration curve in proton dose calculation, while the value obtained for the DirectSPR was higher, 2.8%. The high standard deviations of the multiple irradiation point measurements suggest the presence of residual setup errors affecting the results and the necessity to improve setup repeatability for further investigations. The results of this study confirm the feasibility of the use of DirectSPR and Monoenergetic Plus images generated by the Syngo.via software as input to the Raystation TPS for proton dose calculation. Although preliminary, the results of this study support the reduction of the 3.5% clinical density uncertainty enabled by the implementation of dual-energy CT as suggested by many authors.File | Dimensione | Formato | |
---|---|---|---|
spec_unimi_S65864.pdf
accesso aperto
Descrizione: tesi di specializzazione
Tipologia:
Altro
Dimensione
1.76 MB
Formato
Adobe PDF
|
1.76 MB | Adobe PDF | Visualizza/Apri |
Pubblicazioni consigliate
I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.