In Italy, it is estimated that over a million patients undergo oral anticoagulant treatment, among which Warfarin, commonly known by the trade name Coumadin, is undoubtedly one of the most widely used drugs. The effective administration of anticoagulants aims to inhibit the formation of blood clots while minimizing the risk of bleeding. This demanding and complex task must address numerous disturbances and uncertainties regarding the drug's effects, necessitating regular monitoring and adjustments.This work focuses on assisting and supporting physicians in the intricate management of anticoagulant therapy. A primary contribution involves the development of closed-loop control algorithms, demonstrating improvements in Warfarin therapy compared to reference treatments. Specifically, Proportional-Integral-Derivative (PID) and Model Predictive Control (MPC) controllers were implemented and evaluated in-silico using a well-established pharmacometric model of Warfarin, accounting for inter-subject variability. Various types of references were also assessed in two distinct scenarios, one simpler and one more realistic.The second contribution entails the creation of a decision support application for smartphones using Flutter, facilitating interaction with physicians. The application enables simultaneous experimentation with multiple subjects and tracks International Normalized Ratio (INR) measurements and Warfarin doses taken by patients. Notably, the integration of a PID controller within the app serves as a consultation tool for physicians, suggesting alternative Warfarin doses.
In Italia si stima che oltre un milione di pazienti siano sottoposti a trattamento con anticoagulante orale tra i quali il Warfarin, meglio conosciuto con il nome commerciale Coumadin, è sicuramente uno dei farmaci più utilizzati. La somministrazione efficace di anticoagulanti mira a inibire la formazione di coaguli di sangue riducendo al minimo il rischio di emorragie. Questo compito, così impegnativo e complesso, deve far fronte a numerosi fattori di disturbo e incertezze riguardanti gli effetti del farmaco, che necessita di monitoraggio e aggiustamenti regolari. Questo lavoro si concentra sull'assistenza e il supporto ai medici nella complessa gestione della terapia anticoagulante. Un primo contributo riguarda lo sviluppo di alcuni algoritmi di controllo a circuito chiuso, che hanno dimostrato miglioramenti nella terapia con Warfarin rispetto alle terapie di riferimento. Nello specifico, sono stati creati controllori PID (Proportional-Integral-Derivative) e MPC (Model Predictive Control) e valutati in-silico utilizzando un modello farmacometrico consolidato di Warfarin che tiene conto della variabilità tra soggetti. Sono state inoltre valutate diverse tipologie di "reference" in due scenari distinti, uno più semplice e uno più realistico. Il secondo contributo prevede la creazione di un'applicazione di supporto decisionale per smartphone in Flutter, che consente l'interazione con i medici. L'applicazione consente la sperimentazione simultanea di più soggetti e permette di tener traccia delle misurazioni di INR (International Normalized Ratio) e delle dosi di warfarin assunte dai pazienti. In particolare, l’integrazione di un controller PID all’interno dell’app funge da strumento di consultazione per i medici, suggerendo dosi alternative di Warfarin.
Design of closed-loop algorithms for smart anticoagulant administration and implementation of a decision support App.
BUFFONI, CHIARA MARIA
2023/2024
Abstract
In Italy, it is estimated that over a million patients undergo oral anticoagulant treatment, among which Warfarin, commonly known by the trade name Coumadin, is undoubtedly one of the most widely used drugs. The effective administration of anticoagulants aims to inhibit the formation of blood clots while minimizing the risk of bleeding. This demanding and complex task must address numerous disturbances and uncertainties regarding the drug's effects, necessitating regular monitoring and adjustments.This work focuses on assisting and supporting physicians in the intricate management of anticoagulant therapy. A primary contribution involves the development of closed-loop control algorithms, demonstrating improvements in Warfarin therapy compared to reference treatments. Specifically, Proportional-Integral-Derivative (PID) and Model Predictive Control (MPC) controllers were implemented and evaluated in-silico using a well-established pharmacometric model of Warfarin, accounting for inter-subject variability. Various types of references were also assessed in two distinct scenarios, one simpler and one more realistic.The second contribution entails the creation of a decision support application for smartphones using Flutter, facilitating interaction with physicians. The application enables simultaneous experimentation with multiple subjects and tracks International Normalized Ratio (INR) measurements and Warfarin doses taken by patients. Notably, the integration of a PID controller within the app serves as a consultation tool for physicians, suggesting alternative Warfarin doses.File | Dimensione | Formato | |
---|---|---|---|
Buffoni_ChiaraMaria.pdf
accesso riservato
Dimensione
13.12 MB
Formato
Adobe PDF
|
13.12 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/20.500.12608/62281