This thesis discusses innovative technologies that can contribute to the early diagnosis of different diseases. In the new era of Personalized Medicine, healthcare is tailored to individual patients based on genetic, environmental and lifestyle factors to identify disease susceptibility, guide preventive measures and optimize treatment, especially in fields like oncology. The focus areas include several -omics disciplines, including the quantitative analysis of radiological images, called Radiomics. Shifting on to Artificial Intelligence and Medicine, we will explore the pros and cons of its application in healthcare. In particular, we propose a new methodology using Formal Methods and Radiomics to solve the challenges present in disease diagnosis; the need for a large amount of data, the trustworthiness of the final decision and the complex explainability of the results. We will see Radiomics with different imaging modalities and then how to combine it with Formal Methods for early diagnosis, achieving promising performance, interpretability of results and increased confidence in expert systems. The present methodology is tested on various anatomical parts of the human body, including the brain, breast, shoulder, lung, liver and soft tissues. In summary, the advancements and challenges in Personalized Medicine and the integration of Computer Science can focus on the role of Formal Methods in addressing some of the limitations of Artificial Intelligence. The hope is that, with technological advancements, Personalized Medicine will transform healthcare.
Questa tesi discute le tecnologie innovative che possono contribuire alla diagnosi precoce di diverse malattie. Nella nuova era della medicina personalizzata, l’assistenza sanitaria viene adattata ai singoli pazienti in base a fattori genetici, ambientali e di stile di vita per identificare la suscettibilità alle malattie, guidare le misure preventive e ottimizzare il trattamento, soprattutto in campi come l’oncologia. Le aree di interesse comprendono diverse discipline -omiche, tra cui l’analisi delle immagini radiologiche, chiamata Radiomica. Passando all’Intelligenza Artificiale e alla Medicina, esploreremo i pro e i contro della sua applicazione in ambito sanitario. In particolare, proponiamo una nuova metodologia che utilizza i Metodi Formali e la Radiomica per risolvere le sfide presenti nella diagnosi delle malattie: la necessità di una grande quantità di dati, l’affidabilità della decisione finale e la complessa spiegabilità dei risultati. Vedremo la Radiomica con diverse modalità di imaging e poi come combinarla con i Metodi Formali per la diagnosi precoce, ottenendo prestazioni promettenti, interpretabilità dei risultati e maggiore fiducia nei sistemi esperti. La presente metodologia viene testata su varie parti anatomiche del corpo umano, tra cui cervello, seno, spalla, polmone, fegato e tessuti molli. In sintesi, i progressi e le sfide della medicina personalizzata e l’integrazione dell’informatica possono concentrarsi sul ruolo dei metodi formali nell’affrontare alcune delle limitazioni dell’Intelligenza Artificiale. La speranza `e che, con i progressi tecnologici, la Medicina Personalizzata trasformi l’assistenza sanitaria.
Radiomics and formal methods: a dynamic duo for medicine in pixels
VARRIANO, Giulia
2024
Abstract
This thesis discusses innovative technologies that can contribute to the early diagnosis of different diseases. In the new era of Personalized Medicine, healthcare is tailored to individual patients based on genetic, environmental and lifestyle factors to identify disease susceptibility, guide preventive measures and optimize treatment, especially in fields like oncology. The focus areas include several -omics disciplines, including the quantitative analysis of radiological images, called Radiomics. Shifting on to Artificial Intelligence and Medicine, we will explore the pros and cons of its application in healthcare. In particular, we propose a new methodology using Formal Methods and Radiomics to solve the challenges present in disease diagnosis; the need for a large amount of data, the trustworthiness of the final decision and the complex explainability of the results. We will see Radiomics with different imaging modalities and then how to combine it with Formal Methods for early diagnosis, achieving promising performance, interpretability of results and increased confidence in expert systems. The present methodology is tested on various anatomical parts of the human body, including the brain, breast, shoulder, lung, liver and soft tissues. In summary, the advancements and challenges in Personalized Medicine and the integration of Computer Science can focus on the role of Formal Methods in addressing some of the limitations of Artificial Intelligence. The hope is that, with technological advancements, Personalized Medicine will transform healthcare.File | Dimensione | Formato | |
---|---|---|---|
Tesi_G_Varriano.pdf
accesso aperto
Dimensione
13.1 MB
Formato
Adobe PDF
|
13.1 MB | Adobe PDF | Visualizza/Apri |
I documenti in UNITESI sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
https://hdl.handle.net/20.500.14242/196309
URN:NBN:IT:UNIMOL-196309