The present thesis analyses the evolution of the modern hospital, considering both healthcare services and clinical engineering, as well as the technological infrastructure, which today represents a fundamental element for introducing new therapies and improving the quality of care. The contemporary hospital is increasingly configured as an interconnected technological ecosystem, where digital innovation and information system integration serve as key tools to ensure continuity, safety and precision in clinical practice. In this context, the Spedali Civili di Brescia represent the largest hospital complex in Lombardy and one of the most significant in Italy. The organization, structured into multiple facilities distributed across the territory, boasts an advanced and continuously evolving technological environment that supports clinical, research and innovation activities. This environment constitutes an ideal setting for developing and testing advanced organizational and infrastructural models, such as the one analysed in this work. The thesis specifically explores the application of the hub & spoke organizational model within the Dermatology Unit of the ASST Spedali Civili di Brescia, with the aim of optimizing the management of clinical data and images, enhancing collaboration among network centres and improving early diagnosis of cutaneous melanoma. The hub & spoke model is based on a network structure in which the main centre, called the hub, equipped with specialized expertise and advanced technologies, coordinates several peripheral centres, the spokes, located throughout the territory. This system ensures continuity of care, diagnostic consistency and optimal use of technological and professional resources. In the specific case analysed, the project involved the recent introduction—thanks to a dedicated donation—of next-generation equipment that represents an important step forward in the field of digital dermatology. The central hub, located in the hospital’s main facility, has been equipped with the Vectra WB360 system, a cutting-edge technology for full-body 3D mapping and early detection of suspicious skin lesions. The Vectra integrates artificial intelligence algorithms capable of assisting clinicians in automated image analysis and longitudinal lesion comparison. The peripheral spoke centres, distributed across the hospital network, have been equipped with newly installed high-resolution digital videodermatoscopes, which allow for the acquisition and sharing of high-quality clinical images. The IT integration among the centres is currently being finalized, aiming to build a connected, secure and interoperable clinical network capable of supporting advanced dermatological diagnosis models and territorial collaboration. The thesis focuses particularly on analysing the technological infrastructure that supports the dermatological network, including both the communication links between the hub & spoke centres and the data and image storage system. The term “infrastructure” refers to the coordinated set of hardware components, software systems, communication tools and network protocols that enable the collection, processing, transmission, storage and secure management of clinical data. It plays a central role in ensuring that data are available in real time for clinicians, can be shared across the hub & spoke network and are adequately protected against failures, data loss, or unauthorized access. The analysis highlighted several strengths of the current infrastructure: a high level of cybersecurity ensured by demilitarized zones (DMZs) that isolate acquisition devices during network communication, thereby protecting sensitive information from external attacks or unauthorized access. The network also includes local backup systems for rapid data recovery in case of malfunction or interruptions and a modular, scalable architecture that facilitates the addition of new nodes or devices without compromising overall performance. At the same time, some critical issues have emerged. The reliance on local physical storage, while ensuring a high level of data protection, limits flexibility compared to modern cloud-based solutions, which are increasingly recommended in the healthcare sector for their scalability, accessibility and interoperability. This choice imposes additional constraints on data management and maintenance and may hinder future network expansion. Another issue concerns the lack of integration with the central patient registry, which currently prevents the creation of a unified and coherent patient archive. This limitation restricts the ability to conduct longitudinal analyses, comprehensively monitor clinical follow-ups and fully leverage the collected data for scientific studies and research projects, reducing process efficiency and clinical continuity. Another relevant aspect analysed in the thesis concerns the colour standardization of imaging devices. This activity is crucial to ensure visual and diagnostic consistency of dermatological images, considering that different devices and environmental variables (lighting, acquisition settings) can affect colour rendering and thus clinical interpretation. To this end, a preliminary RGB analysis was conducted to calibrate the colour output of a newly acquired videodermatoscope, comparing it with an existing device and a digital camera identified by clinicians as the gold standard. This technique examines the distribution of red, green, and blue components within digital images; using MATLAB software, the colour values of pixels were extracted to analyse differences and achieve uniformity across images. The initial calibrations yielded positive outcomes, although some variations still need harmonization across different operational contexts. To ensure long-term colour consistency and diagnostic reliability, it will be necessary to define periodic calibration procedures and implement centralized control over image acquisitions. In the future, the project aims to make the network fully operational, enabling clinicians to consult and compare dermatological images from any workstation, while integrating peripheral systems with the hospital’s central IT platforms. Special attention will be given to evaluating the performance of the artificial intelligence algorithms already embedded in the devices: the goal is not only to use them as diagnostic support but also as effective diagnostic tools, always combined by clinical judgment to ensure reliability and alignment with medical evaluations.
Il presente lavoro di tesi analizza l’evoluzione dell’ospedale moderno sia dal punto di vista dei servizi sanitari e dell’ingegneria clinica, sia sotto il profilo del parco tecnologico, oggi elemento fondamentale per introdurre nuove terapie e migliorare la qualità dell’assistenza. L’ospedale contemporaneo si configura sempre più come un ecostistema tecnologico interconnesso, in cui l’innovazione digitale e l’integrazione dei sistemi informativi rappresentano strumenti chiave per garantire continuità, sicurezza e precisione nella pratica clinica. In questo contesto si collocano gli Spedali Civili di Brescia, che rappresentano oggi il più grande complesso ospedaliero della Lombardia e uno tra i più rilevanti a livello nazionale. L’Azienda, articolata in diversi presidi dislocati sul territorio, vanta un parco tecnologico di eccellenza, in continua evoluzione, che supporta attività cliniche, di ricerca e di innovazione. Tale realtà costituisce il contesto ideale per lo sviluppo e la sperimentazione di modelli organizzativi e infrastrutturali avanzati, come quello analizzato nel presente lavoro. L’elaborato approfondisce infatti l’applicazione del modello organizzativo hub & spoke presso l’Unità Operativa di Dermatologia della ASST Spedali Civili di Brescia, con l’obiettivo di ottimizzare la gestione dei dati clinici e delle immagini, potenziare la collaborazione tra i centri della rete e migliorare la diagnosi precoce del melanoma cutaneo. Il modello hub & spoke si basa su una struttura a rete in cui il centro principale denominato hub, dotato di competenze specialistiche e tecnologie avanzate, coordina una serie di centri periferici, gli spoke, dislocati sul territorio. Questa organizzazione consente di garantire continuità assistenziale, uniformità diagnostica e assicurare un uso ottimale delle risorse tecnologiche e professionali. Nel caso specifico in analisi il progetto ha visto la recente introduzione, grazie a una donazione dedicata, di strumentazioni di ultima generazione che rappresentano un importante passo avanti per l’ambito della dermatologia digitale. Il centro hub, collocato nel polo centrale della struttura ospedaliera, è stato dotato del sistema Vectra WB360, una tecnologia all’avanguardia per la mappatura tridimensionale completa del corpo e la rilevazione precoce delle lesioni cutanee sospette. Il Vectra integra algoritmi di intelligenza artificiale capaci di assistere il medico nell’analisi automatizzata delle immagini e nel confronto evolutivo delle lesioni nel tempo. I centri spoke, distribuiti tra i diversi poli territoriali dell’ospedale, invece, sono stati equipaggiati con videodermatoscopi digitali ad alta risoluzione, anch’essi recentemente installati, che consentono l’acquisizione e la condivisione di immagini cliniche di elevata qualità. L’integrazione informatica tra i diversi centri è attualmente in fase di completamento, ma mira a costruire una rete clinica connessa, sicura e interoperabile, in grado di supportare modelli avanzati di diagnosi dermatologica e collaborazione territoriale. Il lavoro di tesi si concentra in particolare sull’analisi dell’infrastruttura tecnologica che supporta la rete dermatologica, comprendente sia la rete di comunicazione tra i centri hub e i centri spoke sia il sistema di archiviazione dei dati e delle immagini. Per infrastruttura si intende l’insieme coordinato di componenti hardware, software, sistemi di comunicazione e protocolli di rete che consentono la raccolta, l’elaborazione, la trasmissione, l’archiviazione e la gestione sicura dei dati clinici. Essa ha un ruolo centrale nel garantire che i dati siano disponibili in tempo reale per i clinici, che possano essere condivisi tra i diversi nodi della rete hub & spoke e che siano adeguatamente protetti da guasti, perdite o accessi non autorizzati. L’indagine ha permesso di mettere in evidenza diversi punti di forza dell’attuale infrastruttura: in primo luogo, un elevato livello di sicurezza informatica, assicurato dalla presenza di zone demilitarizzate (DMZ) che isolano i dispositivi di acquisizione durante la comunicazione in rete, proteggendo così le informazioni sensibili da accessi non autorizzati o attacchi esterni. Inoltre, la rete è dotata di sistemi di backup locali, che consentono il recupero rapido dei dati in caso di malfunzionamenti o interruzioni e di una struttura modulare scalabile, che facilita l’aggiunta di nuovi nodi o di dispositivi senza compromettere le prestazioni complessive. Allo stesso tempo, l’analisi ha evidenziato alcune criticità significative. La dipendenza da un’archiviazione fisica locale, pur garantendo un elevato livello di protezione dei dati clinici, limita la flessibilità rispetto alle moderne soluzioni cloud-based, oggi sempre più raccomandate in ambito sanitario per la loro scalabilità, accessibilità e interoperabilità. Tale scelta comporta anche maggiori vincoli nella gestione e manutenzione dei dati e può rappresentare un ostacolo all’espansione futura della rete. Un’altra criticità riguarda la mancata integrazione con l’anagrafica centrale aziendale, che attualmente impedisce la creazione di un archivio unico e coerente dei pazienti. Questa mancanza limita la possibilità di condurre analisi longitudinali, di monitorare in maniera completa il follow-up clinico e di valorizzare i dati raccolti per studi scientifici e progetti di ricerca, riducendo l’efficienza e la continuità dei processi clinici. Un ulteriore aspetto rilevante approfondito nel lavoro riguarda la standardizzazione cromatica dei dispositivi di acquisizione. Tale attività è risultata cruciale per garantire la coerenza visiva e diagnostica delle immagini dermatologiche, considerando che i diversi dispositivi e le variabili ambientali (illuminazione, impostazioni di acquisizione) possono influenzare la resa cromatica e quindi l’interpretazione clinica. A tal fine, è stata condotta un’analisi RGB preliminare per la calibrazione cromatica di un videodermatoscopio di nuova acquisizione, confrontandolo con un videodermatoscopio già in uso e con una fotocamera digitale considerata dai clinici come gold standard. Questa tecnica consente di esaminare la distribuzione delle componenti di colore (rosso, verde e blu) all’interno delle immagini digitali e, utilizzando il software MATLAB, è stato possibile estrarre i valori cromatici dei pixel, così da analizzare le differenze tra le immagini e procedere alla loro uniformazione. Le calibrazioni iniziali hanno mostrato risultati positivi, tuttavia restano differenze da armonizzare tra i diversi contesti operativi. Per garantire la coerenza cromatica e l’affidabilità diagnostica nel tempo, sarà quindi necessario definire procedure di calibrazione periodiche e implementare controlli centralizzati sulle acquisizioni. In futuro, il progetto mira a rendere la rete completamente operativa, consentendo ai clinici di consultare e confrontare le immagini da qualsiasi postazione e integrando i sistemi periferici con le piattaforme informatiche centrali dell’ospedale. Particolare attenzione sarà dedicata alla valutazione delle prestazioni degli algoritmi di intelligenza artificiale già presenti nei dispositivi: l’obiettivo non è solo impiegarli come supporto alla diagnosi, ma anche come strumento diagnostico effettivo, sempre affiancato al giudizio clinico per garantire affidabilità e coerenza con le valutazioni mediche.
Integrazione di dispositivi dermatologici digitali in ambiente ospedaliero: il caso degli Spedali Civili di Brescia
BRAMBILLA, ARIANNA;Basso, Alessia
2024/2025
Abstract
The present thesis analyses the evolution of the modern hospital, considering both healthcare services and clinical engineering, as well as the technological infrastructure, which today represents a fundamental element for introducing new therapies and improving the quality of care. The contemporary hospital is increasingly configured as an interconnected technological ecosystem, where digital innovation and information system integration serve as key tools to ensure continuity, safety and precision in clinical practice. In this context, the Spedali Civili di Brescia represent the largest hospital complex in Lombardy and one of the most significant in Italy. The organization, structured into multiple facilities distributed across the territory, boasts an advanced and continuously evolving technological environment that supports clinical, research and innovation activities. This environment constitutes an ideal setting for developing and testing advanced organizational and infrastructural models, such as the one analysed in this work. The thesis specifically explores the application of the hub & spoke organizational model within the Dermatology Unit of the ASST Spedali Civili di Brescia, with the aim of optimizing the management of clinical data and images, enhancing collaboration among network centres and improving early diagnosis of cutaneous melanoma. The hub & spoke model is based on a network structure in which the main centre, called the hub, equipped with specialized expertise and advanced technologies, coordinates several peripheral centres, the spokes, located throughout the territory. This system ensures continuity of care, diagnostic consistency and optimal use of technological and professional resources. In the specific case analysed, the project involved the recent introduction—thanks to a dedicated donation—of next-generation equipment that represents an important step forward in the field of digital dermatology. The central hub, located in the hospital’s main facility, has been equipped with the Vectra WB360 system, a cutting-edge technology for full-body 3D mapping and early detection of suspicious skin lesions. The Vectra integrates artificial intelligence algorithms capable of assisting clinicians in automated image analysis and longitudinal lesion comparison. The peripheral spoke centres, distributed across the hospital network, have been equipped with newly installed high-resolution digital videodermatoscopes, which allow for the acquisition and sharing of high-quality clinical images. The IT integration among the centres is currently being finalized, aiming to build a connected, secure and interoperable clinical network capable of supporting advanced dermatological diagnosis models and territorial collaboration. The thesis focuses particularly on analysing the technological infrastructure that supports the dermatological network, including both the communication links between the hub & spoke centres and the data and image storage system. The term “infrastructure” refers to the coordinated set of hardware components, software systems, communication tools and network protocols that enable the collection, processing, transmission, storage and secure management of clinical data. It plays a central role in ensuring that data are available in real time for clinicians, can be shared across the hub & spoke network and are adequately protected against failures, data loss, or unauthorized access. The analysis highlighted several strengths of the current infrastructure: a high level of cybersecurity ensured by demilitarized zones (DMZs) that isolate acquisition devices during network communication, thereby protecting sensitive information from external attacks or unauthorized access. The network also includes local backup systems for rapid data recovery in case of malfunction or interruptions and a modular, scalable architecture that facilitates the addition of new nodes or devices without compromising overall performance. At the same time, some critical issues have emerged. The reliance on local physical storage, while ensuring a high level of data protection, limits flexibility compared to modern cloud-based solutions, which are increasingly recommended in the healthcare sector for their scalability, accessibility and interoperability. This choice imposes additional constraints on data management and maintenance and may hinder future network expansion. Another issue concerns the lack of integration with the central patient registry, which currently prevents the creation of a unified and coherent patient archive. This limitation restricts the ability to conduct longitudinal analyses, comprehensively monitor clinical follow-ups and fully leverage the collected data for scientific studies and research projects, reducing process efficiency and clinical continuity. Another relevant aspect analysed in the thesis concerns the colour standardization of imaging devices. This activity is crucial to ensure visual and diagnostic consistency of dermatological images, considering that different devices and environmental variables (lighting, acquisition settings) can affect colour rendering and thus clinical interpretation. To this end, a preliminary RGB analysis was conducted to calibrate the colour output of a newly acquired videodermatoscope, comparing it with an existing device and a digital camera identified by clinicians as the gold standard. This technique examines the distribution of red, green, and blue components within digital images; using MATLAB software, the colour values of pixels were extracted to analyse differences and achieve uniformity across images. The initial calibrations yielded positive outcomes, although some variations still need harmonization across different operational contexts. To ensure long-term colour consistency and diagnostic reliability, it will be necessary to define periodic calibration procedures and implement centralized control over image acquisitions. In the future, the project aims to make the network fully operational, enabling clinicians to consult and compare dermatological images from any workstation, while integrating peripheral systems with the hospital’s central IT platforms. Special attention will be given to evaluating the performance of the artificial intelligence algorithms already embedded in the devices: the goal is not only to use them as diagnostic support but also as effective diagnostic tools, always combined by clinical judgment to ensure reliability and alignment with medical evaluations.| File | Dimensione | Formato | |
|---|---|---|---|
|
tesi_magistrale_Basso_Brambilla.pdf
accessibile in internet solo dagli utenti autorizzati
Descrizione: tesi
Dimensione
4.08 MB
Formato
Adobe PDF
|
4.08 MB | Adobe PDF | Visualizza/Apri |
|
executive_summary_Basso_Brambilla.pdf
accessibile in internet solo dagli utenti autorizzati
Descrizione: executive summary
Dimensione
647.35 kB
Formato
Adobe PDF
|
647.35 kB | Adobe PDF | Visualizza/Apri |
I documenti in POLITesi sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
https://hdl.handle.net/10589/246151