The ongoing digitalization of healthcare services offers new opportunities to optimize workflows and enhance patient communication. In diagnostic imaging, adequate pre-examination preparation is crucial for procedural success, as misunderstood or incorrectly applied instructions can lead to delays, rescheduling, or cancelled exams. This study aimed to design and preliminarily evaluate an Artificial Intelligence (AI)-based tool to support patient education and preparation for radiological procedures. Following an analysis of the state of the art and key regulatory frameworks governing AI in healthcare, an anonymous survey was administered to radiology professionals to assess perceived critical issues in patient preparation and the acceptability of digital support tools. In parallel, a prototype was developed featuring a pre-recorded AI-generated video avatar driven by a pre-defined, peer-reviewed informational script. Survey results indicate that patient preparation issues are frequently encountered by practitioners, who showed an overall favorable attitude toward introducing the tool while noting potential barriers, including the digital divide. An exploratory economic and organizational analysis was also conducted, comparing the current video avatar with a potential future transition toward an interactive system powered by Large Language Models and Retrieval-Augmented Generation (RAG). Although this remains a preliminary evaluation based primarily on professional perceptions rather than direct patient outcomes, the findings provide an encouraging baseline for future trials and for the progressive, responsible integration of technological innovation into diagnostic imaging pathways.
Il continuo processo di digitalizzazione dei servizi sanitari offre nuove opportunità per ottimizzare i flussi di lavoro e migliorare la comunicazione con il paziente. Nell’ambito della diagnostica per immagini, un’adeguata preparazione agli esami rappresenta un elemento fondamentale per il corretto svolgimento delle procedure, poiché informazioni non comprese o non correttamente applicate possono contribuire a ritardi, riprogrammazioni o mancata esecuzione delle prestazioni. Il presente lavoro ha avuto l’obiettivo di progettare e valutare preliminarmente uno strumento basato sull’Intelligenza Artificiale (IA) a supporto dell’informazione e della preparazione del paziente agli esami radiologici. Dopo l’analisi dello stato dell’arte e dei principali aspetti normativi relativi all’impiego dell’IA in sanità, è stata condotta un’indagine mediante questionario anonimo rivolto a professionisti del settore radiologico, finalizzata a valutare le criticità percepite nella preparazione del paziente e l’accettabilità di strumenti digitali di supporto. Parallelamente, è stato sviluppato un prototipo costituito da un video-avatar preregistrato, generato mediante IA a partire da uno script informativo preventivamente definito e revisionato. I risultati della survey evidenziano come le problematiche associate alla preparazione del paziente siano frequentemente riportate dai professionisti e mostrano un orientamento complessivamente favorevole verso l’introduzione dello strumento, pur rilevando potenziali barriere, tra cui il digital divide. È stata inoltre condotta un’analisi economico-organizzativa esplorativa, confrontando il video-avatar realizzato con una possibile evoluzione futura verso un sistema interattivo basato su modelli linguistici e Retrieval-Augmented Generation (RAG). Pur trattandosi di una valutazione preliminare, basata principalmente sulle percezioni dei professionisti e non ancora su outcome misurati direttamente sui pazienti, i risultati ottenuti rappresentano un presupposto incoraggiante per successive sperimentazioni e per una progressiva e responsabile integrazione dell’innovazione tecnologica nei percorsi della diagnostica per immagini.
Digitalizzazione dei processi in diagnostica per immagini: sviluppo di uno strumento di Intelligenza Artificiale per la preparazione del paziente
FIERRO, FLAVIO
2025/2026
Abstract
The ongoing digitalization of healthcare services offers new opportunities to optimize workflows and enhance patient communication. In diagnostic imaging, adequate pre-examination preparation is crucial for procedural success, as misunderstood or incorrectly applied instructions can lead to delays, rescheduling, or cancelled exams. This study aimed to design and preliminarily evaluate an Artificial Intelligence (AI)-based tool to support patient education and preparation for radiological procedures. Following an analysis of the state of the art and key regulatory frameworks governing AI in healthcare, an anonymous survey was administered to radiology professionals to assess perceived critical issues in patient preparation and the acceptability of digital support tools. In parallel, a prototype was developed featuring a pre-recorded AI-generated video avatar driven by a pre-defined, peer-reviewed informational script. Survey results indicate that patient preparation issues are frequently encountered by practitioners, who showed an overall favorable attitude toward introducing the tool while noting potential barriers, including the digital divide. An exploratory economic and organizational analysis was also conducted, comparing the current video avatar with a potential future transition toward an interactive system powered by Large Language Models and Retrieval-Augmented Generation (RAG). Although this remains a preliminary evaluation based primarily on professional perceptions rather than direct patient outcomes, the findings provide an encouraging baseline for future trials and for the progressive, responsible integration of technological innovation into diagnostic imaging pathways.| File | Dimensione | Formato | |
|---|---|---|---|
|
Fierro_Flavio.pdf
Accesso riservato
Dimensione
1.4 MB
Formato
Adobe PDF
|
1.4 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/20.500.12608/115975