Cogno, Nicolò (2023)
Agent-Based Modelling of Radiation-Induced Lung Injuries.
Technische Universität Darmstadt
doi: 10.26083/tuprints-00026335
Dissertation, Erstveröffentlichung, Verlagsversion
Kurzbeschreibung (Abstract)
Radiotherapy (RT), which nowadays is integrated in more than 50% of the therapies of new cancer patients, involves the use of ionizing radiation (such as photon beams and ions) as a tool to sterilize cancers. However, the lethal doses to be delivered to the tumours are limited by normal tissue complications. Consequently, constraints must be set on the radiation dose and irradiated volume in order to maintain acceptable toxicity levels. An important role in this context is played by computational models that ultimately provide valuable insights useful for tuning the RT parameters. Their use in the biomedical framework has a well-defined pattern: a theoretical model is initially built on the basis of the available in-vitro and/or in-vivo data and implemented in-silico; the model is then altered until a good match between its output and laboratory data is observed and finally used for predictions in the clinical setting. As yet, however, the tolerance doses for the organs at risk are derived from clinical experience and used as inputs for phenomenological Normal-Tissue Complication Probability (NTCP) models that lack a mechanistic description of the underlying phenomena. This thesis describes the implementation of an Agent-Based Model (ABM) that simulates the onset of Radiation-Induced Lung Injuries (RILI) (namely pneumonitis and fibrosis), complications that can occur in the lungs of patients irradiated in the thoracic region. Although relatively common, the risk factors and progression of the RILI, which eventually lead to respiratory failure and death, haven’t been fully elucidated. Here, the capability of the innovative AB modelling approach to improve patient-specific NTCP estimates while attempting to provide insights on the development of RILI is investigated. With respect to the existing dose-volume histogram-based and tissue-architecture approaches, ABMs can take into account not only the patient-specific geometry and tissue-level parameters, but also spatial information on the dose distribution. As a first step, a 3D model of idiopathic pulmonary fibrosis, which resembles the Radiation-Induced Lung Fibrosis (RILF), was implemented using BioDynaMo, an AB simulation framework. The model, whose agents simulate a partial pulmonary acinus, can replicate previous experimental results and assess the appropriateness of the approach for the purpose. The model was subsequently rescaled to represent an alveolar segment at the cell scale that can be damaged locally by external sources. As a surrogate measure of the RILF severity, the RILF Severity Index (RSI) was introduced, derived by combining the loss in the alveolar volume with the increase in the average concentration of the ExtraCellular Matrix (ECM). The RSI showed qualitative agreement with a similar index obtained using data from computational tomographies and the ECM patterns matched clinical findings. Finally, a pipeline was established that links TOPAS-nBio, a particle transport simulator for biological applications, with BioDynaMo. The alveolar segment structure was rebuilt using TOPAS-nBio and the delivery of realistic dose distributions at the cell scale was simulated. The output was then used as an input for the AB model and the effect of different fractionation schemes and radiation qualities on the outcome explored. In accordance with previous studies, a 5-fractions treatment resulted in a lower RSI with respect to the delivery of the same dose in a single fraction and an increased sensitivity to peaked protons dose distributions with respect to flatter ones from photons irradiation was observed. Overall, the results presented in this thesis prove the capability of the AB models to recapitulate some main radiobiological processes and advise for their potential complementary role in NTCP estimates.
Typ des Eintrags: | Dissertation | ||||
---|---|---|---|---|---|
Erschienen: | 2023 | ||||
Autor(en): | Cogno, Nicolò | ||||
Art des Eintrags: | Erstveröffentlichung | ||||
Titel: | Agent-Based Modelling of Radiation-Induced Lung Injuries | ||||
Sprache: | Englisch | ||||
Referenten: | Durante, Prof. Dr. Marco ; Hamacher, Prof. Dr. Kay | ||||
Publikationsjahr: | 20 November 2023 | ||||
Ort: | Darmstadt | ||||
Kollation: | xiii, 114 Seiten | ||||
Datum der mündlichen Prüfung: | 15 November 2023 | ||||
DOI: | 10.26083/tuprints-00026335 | ||||
URL / URN: | https://tuprints.ulb.tu-darmstadt.de/26335 | ||||
Kurzbeschreibung (Abstract): | Radiotherapy (RT), which nowadays is integrated in more than 50% of the therapies of new cancer patients, involves the use of ionizing radiation (such as photon beams and ions) as a tool to sterilize cancers. However, the lethal doses to be delivered to the tumours are limited by normal tissue complications. Consequently, constraints must be set on the radiation dose and irradiated volume in order to maintain acceptable toxicity levels. An important role in this context is played by computational models that ultimately provide valuable insights useful for tuning the RT parameters. Their use in the biomedical framework has a well-defined pattern: a theoretical model is initially built on the basis of the available in-vitro and/or in-vivo data and implemented in-silico; the model is then altered until a good match between its output and laboratory data is observed and finally used for predictions in the clinical setting. As yet, however, the tolerance doses for the organs at risk are derived from clinical experience and used as inputs for phenomenological Normal-Tissue Complication Probability (NTCP) models that lack a mechanistic description of the underlying phenomena. This thesis describes the implementation of an Agent-Based Model (ABM) that simulates the onset of Radiation-Induced Lung Injuries (RILI) (namely pneumonitis and fibrosis), complications that can occur in the lungs of patients irradiated in the thoracic region. Although relatively common, the risk factors and progression of the RILI, which eventually lead to respiratory failure and death, haven’t been fully elucidated. Here, the capability of the innovative AB modelling approach to improve patient-specific NTCP estimates while attempting to provide insights on the development of RILI is investigated. With respect to the existing dose-volume histogram-based and tissue-architecture approaches, ABMs can take into account not only the patient-specific geometry and tissue-level parameters, but also spatial information on the dose distribution. As a first step, a 3D model of idiopathic pulmonary fibrosis, which resembles the Radiation-Induced Lung Fibrosis (RILF), was implemented using BioDynaMo, an AB simulation framework. The model, whose agents simulate a partial pulmonary acinus, can replicate previous experimental results and assess the appropriateness of the approach for the purpose. The model was subsequently rescaled to represent an alveolar segment at the cell scale that can be damaged locally by external sources. As a surrogate measure of the RILF severity, the RILF Severity Index (RSI) was introduced, derived by combining the loss in the alveolar volume with the increase in the average concentration of the ExtraCellular Matrix (ECM). The RSI showed qualitative agreement with a similar index obtained using data from computational tomographies and the ECM patterns matched clinical findings. Finally, a pipeline was established that links TOPAS-nBio, a particle transport simulator for biological applications, with BioDynaMo. The alveolar segment structure was rebuilt using TOPAS-nBio and the delivery of realistic dose distributions at the cell scale was simulated. The output was then used as an input for the AB model and the effect of different fractionation schemes and radiation qualities on the outcome explored. In accordance with previous studies, a 5-fractions treatment resulted in a lower RSI with respect to the delivery of the same dose in a single fraction and an increased sensitivity to peaked protons dose distributions with respect to flatter ones from photons irradiation was observed. Overall, the results presented in this thesis prove the capability of the AB models to recapitulate some main radiobiological processes and advise for their potential complementary role in NTCP estimates. |
||||
Alternatives oder übersetztes Abstract: |
|
||||
Status: | Verlagsversion | ||||
URN: | urn:nbn:de:tuda-tuprints-263354 | ||||
Sachgruppe der Dewey Dezimalklassifikatin (DDC): | 500 Naturwissenschaften und Mathematik > 530 Physik | ||||
Fachbereich(e)/-gebiet(e): | 05 Fachbereich Physik 05 Fachbereich Physik > Institut für Physik Kondensierter Materie (IPKM) 05 Fachbereich Physik > Institut für Physik Kondensierter Materie (IPKM) > Biophysik |
||||
Hinterlegungsdatum: | 20 Nov 2023 14:42 | ||||
Letzte Änderung: | 21 Nov 2023 09:44 | ||||
PPN: | |||||
Referenten: | Durante, Prof. Dr. Marco ; Hamacher, Prof. Dr. Kay | ||||
Datum der mündlichen Prüfung / Verteidigung / mdl. Prüfung: | 15 November 2023 | ||||
Export: | |||||
Suche nach Titel in: | TUfind oder in Google |
Frage zum Eintrag |
Optionen (nur für Redakteure)
Redaktionelle Details anzeigen |