O3-DPACS Open-Source Image-Data Manager/Archiver and HDW2 Image-Data Display: An IHE-compliant project pushing the e-health integration in the world

Transcription

1 Computerized Medical Imaging and Graphics 30 (2006) O3-DPACS Open-Source Image-Data Manager/Archiver and HDW2 Image-Data Display: An IHE-compliant project pushing the e-health integration in the world Paolo Inchingolo a,b,, Marco Beltrame a,b, Pierpaolo Bosazzi a,c, Davide Cicuta a, Giorgio Faustini a,b, Stefano Mininel a,b, Andrea Poli a,b, Federica Vatta a,b a Health Telematics Lab (HTL), Open Source Lab (OSL) & Neuromodelling Lab (NML), DEEI, University of Trieste, Italy b Higher Education in Clinical Engineering (SSIC-HECE), University of Trieste, Italy c Clinical Unit of Radiology, University of Trieste, Italy Abstract After many years of study, development and experimentation of open PACS and Image workstation solutions including management of medical data and signals (DPACS project), the research and development at the University of Trieste have recently been directed towards Java-based, IHE compliant and multi-purpose servers and clients. In this paper an original Image-Data Manager/Archiver (O3-DPACS) and a universal Image-Data Display (HDW2) are described. O3-DPACS is also part of a new project called Open Three (O3) Consortium, promoting Open Source adoption in e-health at European and world-wide levels. This project aims to give a contribution to the development of e-health through the study of Healthcare Information Systems and the contemporary proposal of new concepts, designs and solutions for the management of health data in an integrated environment: hospitals, Regional Health Information Organizations and citizens (home-care, mobile-care and ambient assisted living) Elsevier Ltd. All rights reserved. 1. Introduction Picture Archiving and Communication Systems (PACSs) are often the heart of modern information systems in radiology departments [1]. Nowadays, a PACS usually identifies a manager and archiver of images running on dedicated hardware which, despite the use of standards such as DICOM (Digital Imaging and COmmunications in Medicine) [2 4], is to some extent forced to proprietary expensive and hardly scalable and expandable choices. This situation naturally makes the conditions on what the hospital should use around the PACS, such as workstations and modalities, since these are supposed to work keeping with the PACS [1,5,6]. Corresponding author. addresses: [email protected] (P. Inchingolo), [email protected] (M. Beltrame), [email protected] (P. Bosazzi), [email protected] (D. Cicuta), [email protected] (G. Faustini), [email protected] (S. Mininel), [email protected] (A. Poli), [email protected] (F. Vatta). Although PACSs were invented in the early 1980s, still in 1990 Freiherr described them as a premature paradise [7]. Even the few PACS installations that gained benefits in the 1990s were stuck with the problem of the high costs, related especially to keeping updated and working the systems in an environment in which the lack or an incomplete adoption of standards meant a nearly-every-day work for patching, tuning and updating software, to make sure that they can cooperate with new medical instrumentation or new third-party software [1,5,8 11]. This issue, in the most of cases connected to the limited expansibility of PACS systems, led to the home-made practice, hence implying the development of custom add-ons and self-made patches [12 14]. The growth of this specialized and customized software put a lot of obstacles toward the concept of integration, leading oppositely the market to proprietary interpretation of standards rather than to a definition of a common interface. In 2001, in a review of the Next-generation Health Care published by the World Medical Association, Inchingolo [15] underlined that many key aspects were still to be solved for PACS success. As only in-house big projects as, e.g., Papyrus [16], reached real and stable results, the author proposed a Deca /$ see front matter 2006 Elsevier Ltd. All rights reserved. doi: /j.compmedimag

2 392 P. Inchingolo et al. / Computerized Medical Imaging and Graphics 30 (2006) logue of rules to be fulfilled by Healthcare Enterprise in PACS implementation and management. Still today, although the penetration of PACSs increased considerably in the last 4 years, one of the main issues to consider is the integration between RIS (Radiology Information System) and PACS [13,14,17 20], which was one of main aims of DICOM and HL7 supporters. This issue should be considered as a transitional step towards the realization of a new generation of hospital information systems based on the integration of distinct elements, in opposition to the development of monolithically built software with the features of both systems, lacking of the fundamental characteristic of upgradeability and real e- health integration. The way to reach this goal is to integrate systems basing not on data typology, as used in last years, but on the role they assume in the information workflow. The Higher Education in Clinical Engineering (HECE) Program and the Health Telematics Laboratory (HTL) of the Bioengineering and ICT Group at DEEI, University of Trieste, have been working on these themes for a long time (since 1991). The main results obtained in the last 4 years are described and explained in this paper. The recent results and the strategies of HTL in this field have been significantly conditioned by the two last EuroPACS meetings, hold in Oulu (Finland) in 2002 and in Trieste (Italy) in The presentation, as concluding lecture of the EuroPACS meeting in Oulu in 2002, of the new trends of the DPACS (Data and Picture Archiving and Communication System) project at the University of Trieste [21] introduced the DPACS advances to foster the e-health integration in the enlarged Europe and the combined hospital territory citizen environment. The recognized importance of these strategies for the future of Europe led the EuroPACS Society to entrust HECE with the organization of the 2004 EuroPACS meeting in Trieste, focusing on these themes. The successful EuroPACS-MIR 2004 in the enlarged Europe meeting in September 2004 [22], with more than 400 participants from 47 Countries, witnessed the deep discussion on the organizational, standard-related and interoperability issues in all the contexts from the single department case up to the transnational integration. The real problems of managing the administrative and clinical workflows in any complex situation, including hospital territory citizen integration and multi-lingual management, have been also deeply discussed and experimented through a living-lab that has been built-up in the Conference exhibition area, connecting most of the exhibited PACS and RIS systems as well as some hospitals in Italy and in the surrounding Countries. The specific problems related to the transitional and developing Countries have been also addressed. Discussions in all the conference sessions, and especially the ones on interoperability [4,23 25] in the 1-day-lasting workshop on the world-wide IHE (Integrating the Healthcare Enterprise) project, gave strong results and guidelines for the future work. First, the round table Is there a need for a transnational IHE committee in Central and Eastern Europe? concluding the IHE Workshop closed with the commitment to HECE of creating a transnational IHE committee for the Central and Eastern Europe, dealing with technical, harmonization and law-orienting activities in 22 Central and Eastern European Countries. Second, the same round table and most of the IHE workshop sessions underlined that the adoption of open standards and open source solutions is becoming a strictly obligated path to facilitate a fast integration of health systems in Europe and world-wide, fostering this process in the transitional and developing Countries [26]. In fact, only well supported open source solutions can ensure a world-wide adoption of manager/archiver and cross-enterprise document sharing systems for medical data and images; furthermore, open source promotes the expandability and modularity of solutions, thus allowing collaboration among developers [27 38]. Finally, open-source assures stability in the service use of the products, being the user the health-care system in the general sense able to use it with long-term continuity and also to improve, update, modify and integrate it as much as necessary for his scopes, even in case the developer or the vendor disappears from the market. HECE, together with HTL and OSL (Open Source Laboratory) at DEEI, started in 2005 both these lines. In particular, in relation to the second one, the group of Trieste, who presented the new open-source version of their DPACS-2004 project [39] together with a universal workstation named HDW2 [40], and the group of the Radiology Department of Padova, which presented the new open-source version of their Raynux/MARiS project [41], decided to fuse and integrate their projects and efforts. Hence, the Open Three (O3) Consortium Project has been started [42]. O3 deals with the three domains of the tomorrow s e-health, in the frame of the European e-health programs: hospital, territory and home-care/mobile-care/ambient assisted living (AAL). 2. Methods 2.1. Premises and the DPACS project The research work at the University of Trieste carried out by HECE and HTL on PACSs started nearly 15 years ago, in 1991, after that a CommView AT&T Philips multi-site PACS system was installed in 1988 at the Trieste s Hospitals Cattinara and Maggiore. This was the first European installation of a commercial PACS system in a hospital enterprise and also the first installation in Europe of two PACS systems connected together over a metropolitan area network. Despite being this system so innovative for that time and the very positive results obtained with the re-organization of the Radiology Department work [43], it was early clear that the proprietary installation was forcing users to adapt to the system and not vice-versa. Some issues became evident: high running costs, unaffordable cost of upgrade, difficulty in creating large archives, low performance, no possibility to integrate non- CommView PACS systems and to let them share data, forced use of expensive customized workstations for data visualization and no easy access from outside the hospitals. The HTL work aimed to overcome these limitations of the Commview PACS System and to open the proprietary installation by developing versatile and open source tools (essentially gateways and client workstations) for LAN, MAN and WAN

3 P. Inchingolo et al. / Computerized Medical Imaging and Graphics 30 (2006) communications with the PACS [44]. By this way, it has been possible distributing images over the hospital departments and surgery rooms of the three hospitals and the bioengineering and medical physics research centres of Trieste, with some connections overseas to the National Institutes of Health at Bethesda, MD (USA), stimulating also the growth of the Informative Trieste System [45]. In 1994, the first PACS browsing interface was developed, allowing virtually world-wide images distribution without dedicated client software [46]. However, as the research, the related results and the clinical experimentation proceeded, it became clear that an impassable limit was reached, due to the intrinsic limitations of the Commview PACS System. For this reason, in 1995 the project of a totally new system named DPACS (Data and Picture Archiving and Communication System) has been started [47,48]. The goal of DPACS was the development of an open, scalable, cheap and universal system with accompanying tools, to store, exchange and retrieve all health information of each citizen at hospital, metropolitan, regional, national and European levels, thus offering an integrated virtual health card of the European Citizens. A first version of DPACS was experimented in at the Cattinara Hospital [49]. In 1998 the DPACS system was running routinely for managing all radiological images (CT, MRI, DR, US, etc.) as well as in the connection with the stereo-tactic neurosurgery, thus substituting completely the old AT&T Commview PACS system. Some mono-dimensional signals such as ECGs were also integrated into the system. The philosophy of this version of the DPACS project is reported in Fig. 1. The DPACS project highly favoured the cohesion of Trieste s scientific and key-service activities and triggered the start and realization in 2000 of a new large project, called the New Telematic Network of 2000 s Trieste, carrying also many types of activities, from e-learning to telemedicine [50]. Over the years, DPACS was enriched with the sections of anatomo-pathology, anaesthesia and reanimation, clinical chemistry laboratory and others; furthermore, its application has been progressively forwarded to the new emerging necessities of the future health care and assistance to the world citizen, based on telemedicine-driven home-care, personal-care and ambient assisted living. Consequently, some new needs have been pointed out and used to program the new developments of the project, as: To have a multi-lingual approach to both client and server managing interfaces and to the presentation of medical contents [51,52]. To have a Simple Data & Image Display client interface, automatically updatable, highly portable from a PC or a MAC or a LINUX workstation to a palm or a cellular-based communicator [53 58]. To be able to connect with a wide variety of communication means, both fix and mobile [59,60]. To offer a highly modular data &, independent of the platform (UNIX/LINUX, WINDOWS, MAC) and of the selected data-base. To improve the interoperability of both server and client system components among them and with all the other information systems components in the hospital and in the health enterprise [8,23]. To solve all the security problems. To have an efficient and effective tool to create the integrated virtual clinical record in the hospital as well as at home or during the travel of a citizen [58,61] New trends and new strategic choices of the DPACS project The first version of the flexible Java-based workstation, called HDW (HTL DICOM Workstation), developed from 2000 to 2002 [21] has been the first result to give a solution to most of the above listed needs for the client interface [62 65]. HDW1, a very versatile platform-independent image-data display workstation, based on Java, providing user authentication and easy use also for terminal (private) users, was able to retrieve, visualize and manage medical images and also various other types of clinical data, as ECGs and pneumological data. HDW1 was studied, developed and validated in Regione Friuli Venezia Giulia Fig. 1. The philosophy of the DPACS project, Version 1997, installed and running routinely at the Cattinara hospital from 1998 to mid 2005.

4 394 P. Inchingolo et al. / Computerized Medical Imaging and Graphics 30 (2006) with the contribution of Insiel spa, the regional public company managing all the information systems of the Region in health and public administration. The strong feedback taken from radiologists and physicians of other specialities where HDW1 was installed has been the base for the further development of HDW and also for the settling up of a totally new project for the DPACS Image-Data Manager/Archiver (see below). The new HDW, called HDW2 [40] and the new DPACS, called DPACS-2004 [39] and re-named O3-DPACS after the constitution of the O3 Consortium, have several key common features related to the technical, methodological and organizational domains but one diversity, related to the market aspect. The common features are the higher scalability and modularity, the use of Java and related software at any level, the support of any platform, the interoperability through the full compliance to the Integrating the Healthcare Enterprise [IHE] [66] world project, the high level of multi-language internationalization (interface and content), the management of any type of medical information, i.e., images, data and signals, the high level of security and safety management, the support of various types of data-bases and application context, embracing the three domains of the tomorrow s e-health in the frame of the European e-health programs: hospital, territory and home-care/mobile-care /AAL. The diversity is due to the fact that, as the development of HDW has been co-supported by Insiel, it is currently distributed as a priced licensed product and the software source code is therefore not available to the users. This solution has been chosen particularly for market and economical common decisions. On the other side, DPACS-2004/O3-DPACS has been thought since the first phases of development as an open-source product, and it is distributed (with the source code) with a GNU GPL license. Among the many improvements of DPACS-2004/O3- DPACS with respect to the original DPACS systems running in the hospitals of Trieste since 1996, a relevant one is dealing with the operating system (OS) and the data-base management system (DBMS). The first PACS software was running only on a Unix platform and with the Oracle DBMS; therefore, it was platform-dependent (essentially OS dependent) as well as dependent on the DBMS. The growth of Java as a stable platform [17,63,67,68] made this solution a valid candidate to build up a cross-platform O3-DPACS server, working either with Oracle DBMS or with an open source DBMS, like PostgreSQL, as well as with other products, as long as proper JDBC (Java Database Connectivity) drivers are available. Furthermore, the availability of standard interfaces to different DBMSs allowed designing the new O3-DPACS as a system able to run either on Microsoft Windows or on Linux operating systems. The same choice was taken by the HDW2: also in this case, thanks to the Java choice, there are no constraints on the operating system: the workstation has been successfully tested for Windows, Linux and Mac OsX environments. Furthermore, taking advantage of PACS supporting activity, the administration tools included in O3-DPACS are being continuously improved to give response to the real needs of operators in radiology departments or in any other medical facility. The last but probably most important issue of O3-DPACS is interoperability. The first DPACS had to face a plethora of problems to connect and communicate with other equipments even inside the Cattinara hospital, since no guidelines and no homogeneous use of standards were available at that time. One of the best efforts has been working on the information flow in order to overcome these barriers and in the last years the birth and growth of IHE (Integrating the Healthcare Enterprise) started bringing order in communication among systems, allowing interoperability on wide scale and among different brands. As a consequence of these new trends, several IHE Integration Profiles on both O3-DPACS and HDW2 have been implemented and tested for interoperability at the 2005 European IHE Connectathon held in Nordwijkerhoud (The Netherlands); many other new integration profiles have been implemented after the 2005 Connectathon and tested recently for interoperability at the 2006 European Connectathon held in Barcelona (Spain) The O3-DPACS Open-Source Image-Data Manager/Archiver O3-DPACS is a Java J2EE (Java 2 Enterprise Edition) application, developed on the open source application server JBOSS [69]; in addition, it can run on nearly every application server as the concept of platform independence states. It has been realized as a modular collection of services [70], as summarized in Fig. 2. As communication protocols, DICOM is used mainly for clinical data, signals and images and HL7 (Health Level 7) [71] for administrative data. In the left side of Fig. 2, the DICOM and HL7 communication modules connect the external world with the modules of the services: Storage: to store DICOM objects. The server allows the storage of: (1) images (as all the PACS systems), (2) data, such as reports, in the form of DICOM Structured Reports, Presentation States, and (3) waveforms, such as ECG, EMG, EEG, etc., as DICOM Waveform. Query/retrieve of DICOM objects: to execute query and retrieval of the stored objects via DICOM protocol. Modality Performed Procedure Step: to receive messages about the exam completion status and link them to the stored data. DICOM Storage Commitment: to verify that data are properly stored and to confirm this to modalities. HL7 Message Interpretation: to manage administrative data, as identifying information exchange, or check for realignment of not consistent patient information, i.e., due to a first-aid procedure. The system has been designed to work inside and outside IHE environments. According to the IHE Actors definition, O3-DPACS, in the Version of May 2005 (Nordwijkerhoud Connectathon), has been developed and implemented as the combination of (Table 1):

5 P. Inchingolo et al. / Computerized Medical Imaging and Graphics 30 (2006) Fig. 2. The modular structure of O3-DPACS. (1) an, the actor that archives and manages radiological images and added objects allowing queries and several kinds of access to information. (2) a Data Manager/Archiver, an actor that analogously archives and manages clinical data/waveforms and added objects. This second actor is currently defined only for some IHE domains, as for Cardiology, while for most of the others IHE domains work is in progress or it has still to be started. Moreover, some other actors have been developed and implemented (see Table 1): (3) a Performed Procedure Step Manager, that manages messages from modalities, indicating state of advancement in an examination. Table 1 O3-DPACS and HDW2 IHE actors O3-DPACS Audit Record Repository Imaging Document Source Performed Procedure Step Manager Secure Node Time Client HDW2 Document Consumer Evidence Creator Image Display Imaging Document Consumer Portable Media Creator Print Composer Report Creator/Reader Secure Node Time Client (4) an Audit Record Repository, that archives and manages the messages recording the events of all actors present in a configuration. (5) a Secure Node, that provides system authentication and access control of user and of any connected node. (6) a Time Client, that synchronizes with a trusted remote clock. All the main IHE integration profiles necessary to work with these actors have been developed and implemented (Table 2): ARI: Access to Radiology Information, applied to the Image Manager/Archiver actor, for specifying access modality to DICOM data, so that they can be found and retrieved in a coherent way. CPI: Consistent Presentation of Images, applied to the Image Manager/Archiver, to manage the Presentation States, i.e., the objects that state how radiological images are to be viewed. CT: Consistent Time, applied to the Time Client actor, to synchronize time for all the enterprise-wide systems. PIR: Patient Information Reconciliation, applied to the Image Manager/Archiver and the Performed Procedure Step Manager actors, to solve some common problems about patient registration in the hospital enterprise. SEC: Basic Security, applied to the, the Performed Procedure Step Manager, the Secure Node and the Audit Record Repository actors, to manage secure communication via TLS (Transport Layer Security Charter) and to perform extensive logging of the related operations. SWF: Scheduled Workflow, applied to and Performed Procedure Step Manager actors, to assure cooperation in a real radiology workflow. These IHE actors and integration profiles assure all the complete interoperability of O3-DPACS as a PACS in a safe, secure

6 396 P. Inchingolo et al. / Computerized Medical Imaging and Graphics 30 (2006) Table 2 O3-DPACS and HDW2 IHE compliance matrixes IHE profile O3-DPACS Access to Radiology Information (ARI) Audit Trail Node Authentication (ATNA) Basic Security (SEC) Consistent Presentation of Images (CPI) Consistent Time (CT) Cross-Enterprise Document Sharing for Imaging (XDS-I) Evidence Documents (ED) Key Image Notes (KIN) Patient Information Reconciliation (PIR) Reporting Workflow (RWF) Scheduled Workflow (SWF) HDW2 Audit Trail Node Authentication (ATNA) Basic Security (SEC) Access to Radiology Information (ARI) Consistent Presentation of Images (CPI) Consistent Time (CT) Cross-Enterprise Document Sharing (XDS) Cross-Enterprise Document Sharing for Imaging (XDS-I) Key Image Notes (KIN) Portable Data for Imaging (PDI) Reporting Workflow (RWF) Scheduled Workflow (SWF) Actor Secure Node Performed Procedure Step Manager Secure Node Audit Record Repository Time Client Imaging Document Source Performed Procedure Step Manager Performed Procedure Step Manager Performed Procedure Step Manager Secure Node Image Display Secure Node Image Display Report Creator/Reader Image Display Print Composer Evidence Creator Time Client Document Consumer Document Consumer Imaging Document Consumer Image Display Evidence Creator Portable Media Creator Image Display Print Composer Report Creator/Reader Report Creator/Reader Image Display Evidence Creator For each implemented profile the actors using it are listed. an integrated environment and with the capability to solve multiple registrations of a same patient. In the last year of research, the main interest has been directed to two new, emerging aspects: (1) the organization of flows and actions of medical reporting and (2) the exchange of documents and images across healthcare enterprises. To achieve the first goal, the following new integration profiles have been added to the actor: ED: Evidence Documents, for optimal management within the radiological workflow of all added objects created in referral steps (report, notes, graphic objects). KIN: Key Image Notes, to create notes on details of some images to make them more recognizable, as well as to apply and visualize them in a simple and intuitive way. RWF: Reporting Workflow, to allow any step in the complex referral procedure to be monitored and organized so as to guarantee maximum simplicity and transparency, which are fundamental requisites to obtain a complete, true and errorfree diagnosis. The Reporting Workflow profile has been applied also to the Performed Procedures Step Manager actor, to assure the tracing

7 P. Inchingolo et al. / Computerized Medical Imaging and Graphics 30 (2006) of the state of advancement of the Reporting Workflow relative to an examination. To achieve the second goal, i.e., the exchange of documents and images across healthcare enterprises, a new actor called Imaging Document Source has been implemented, and the following integration profile has been associated to it: XDS-I: Cross-Enterprise Document Sharing for Imaging, which allows retrieving the health enterprise original clinical images from outside in any moment. Finally, since the assessment of the SEC profile has been proved its weakness in assuring safety and security in composite and heterogeneous health systems, a new profile for the Secure Node actor has been defined and subsequently developed on O3-DPACS: ATNA: Audit Trail Node Authentication, which manages security in all communications between heterogeneous systems. The development of most of these new profiles and of the new Imaging Document Source actor has required a preliminary research for finding and setting up the most suitable solution and interpretation of the profiles themselves, with much cooperative work with other research centres at international level. More information about the O3-DPACS IHE Integration Statements and DICOM Conformance Statements can be found in the O3-DPACS web pages [72] and in the official IHE website [66]. All the O3 modules are downloadable directly from the O3 Consortium web-site [73]. Being O3-DPACS currently installed and working routinely in the hospitals of Trieste, Padova and Pisa, its deep testing in these three large sites gives continuous feedback, which is used to improve some functions and, in particular, to add new additional features. The primary purpose of these new tools is to help in the daily routine of radiology and also other departments, in order to avoid the trouble of lack of standardization and interoperability or wrong use of instruments. This is what is called bad habit, such as typing wrong names or wrong codes. Hence, modules to move studies from one patient to another one or to face non-uniqueness of DICOM instances or lack of patient identifiers in the hospital enterprise have been developed, to allow working also in a non-ihe environment in order to facilitate the transition to interoperability in the environment itself. These modules are shown in the right side of Fig. 2, where Web access components are pictured. The first ones are Management Modules written as JSP (Java Server Pages), which are optimized to allow the operators to perform some routine tasks on O3- DPACS and on its database, and some other JSP pages optimized to allow remote assistance for more serious troubles, like some of the above discussed bad habits. They have been designed to run on a JBOSS internal Apache Tomcat web server [74] with configurable port (e.g. 80, 8080); this solution allows also avoiding firewall complications frequently encountered in healthcare environments. Moreover, a web-services access module allows applications and browsers to retrieve medical data and images virtually from any internet-enabled device, including mobiledevices [75]. The web service communication type is required by the IHE ITI (Information Technology Infrastructure) Integration Profiles, which are under development and are going to be tested to satisfy the need for broad access to medical information. Finally, an auto-routing module has been designed and implemented to deliver images to other destinations according to some configurable attributes such as DICOM Calling AE-Title (Application Entity Title), or the data type, or many attributes from DICOM list. As a matter of fact, O3-DPACS scalability and modularity were thought also to make DPACS easily usable in the territorial environment, and basically to make it adaptable from the needs of a big hospital to those of a small clinic as well as from secure intra-net intra-hospital use to secure access from patient s home or physician s ambulatory. Finally, a powerful module developed by O3-PD, the section of Padova of the O3 Consortium, permits a very efficient backup of the data on multiple media The HDW2 Image-Data Display The HDW2 Workstation has the functionalities of a display of both images and data/signals. Its key features are modularity and configurability: it is possible to manage the application in order to let it fit to every kind of environment; it is also possible to configure every module, to enable/disable some particular functionality or to enhance the authentication procedure. HDW2 fully supports authentication. Three different authentication systems are provided: the smart card authentication, a basic username/password authentication or a network authentication; the last one provides a fully traceable sub-system for the management of the use of HDW2. Fig. 3 shows the object-oriented design of HDW2: there is a common base of essential services for the correct management of different types of data for different source types: it is possible to open medical files from a remote image/data manager (from a PACS) or from the local disk; the file can contain images, signals or pure SR (Structured Reporting) data. In a deep analysis, the system can be seen in this way: authentication and logging systems are responsible for security issues; configuration manager together with the application workflow management take care of the correct way of the procedure being working. Once the clinical data are taken by one of the different sources as a PACS, a DICOMDIR (a unique and mandatory DICOM File within a File Set which contains the Media Storage Directory SOP Class) or a local disk, they can be viewed using one the appropriate viewers: Image Viewer (also for multi-slice images); DICOM Structured Reporting (SR) viewer; Waveform Viewer.

8 398 P. Inchingolo et al. / Computerized Medical Imaging and Graphics 30 (2006) Fig. 3. Object-oriented design of HDW2. It is possible to work on these types of data using also other modules: Print Module, allowing a DICOM and non-dicom print procedures; Study Move Module, letting the user moving the selected data from a PACS server to another one; Patient CD module, providing a simple way to burn exams on removable devices, like CDs or DVDs, also creating an html tree model with all the images on jpeg format. DICOM SR is used both to present numerical data as those related to the Clinical Chemistry Laboratory, as well as to offer a new type of presentation of medical reports, based on structuring the report contents. This approach has been started with radiological reports, and it is under test and evaluation within an international multi-centric group of radiologists. HDW2 supports many add-on modules, with either commercial or GNU GPL (open source) license. One of the most claimed modules is O3-3D, an open-source module for the three-dimensional reconstruction and manipulation of images, designed at the beginning of 2006; its version O3-3Dv1 is scheduled to be fully developed and tested before the end of the year. O3-3D can fully interoperate with HDW2 as well as with any other workstation being or to be developed within or outside O3. O3-3D benefits of a long tradition of the group of Trieste in 3D reconstruction (and related 3D rendering) of brain electrical activity in pathological conditions, starting from MRI images and EEG maps project TEBAM [76 79]. Another add-on module is O3-TEBAM, which actually performs, using an 8 16 CPUs computing system, the 3D electrical brain activity reconstruction [79]. A third one is O3-MIRC which permits to connect HDW2 with any Medical Imaging Resource Centre (MIRC) site or with a local MIRC one, allowing the radiologist or another medical specialist or a teacher to retrieve MIRC images, in order to compare them with his/her images under examination or for teaching purposes. According to the IHE Actors definition, HDW2, in the version of May 2005 (Nordwijkerhoud Connectathon), has been developed and implemented as (Table 1): (1) An, the actor that archives and manages radiological images and added objects allowing queries and several kinds of access to information. (2) A Secure Node (see above point (5) for its characteristics). (3) A Time Client (see above point (6) for its characteristics). As for the case of O3-DPACS, also for HDW2 the Consistent Time profile (CT) has been applied to the Time Client and the Basic Security profile (SEC) has been applied to the Secure Node actor in the 2005 version, while the new Audit Train Node Authentication profile has been implemented to the Secure Node actor during the last year. Four main IHE integration profiles necessary to work with the have been developed and implemented (Table 2): ARI: Access to Radiology Information; CPI: Consistent Presentation of Images; SEC: Basic Security; SWF: Scheduled Workflow. Analogously to O3-DPACS, also HDW2 has been opened during the last year to the new frontiers (1) of the organization of reporting information flows and (2) of the exchange (retrieval) of documents and images across healthcare enterprises. To achieve the first goal, five new actors have been introduced and implemented in HDW2: (4) Evidence Creator: to create added objects on the image, such as notes or graphic references. (5) Report Creator: to produce electronic structured reports. (6) Report Reader: to read electronic structured reports.

9 P. Inchingolo et al. / Computerized Medical Imaging and Graphics 30 (2006) The following integration profiles have been associated to these new actors: KIN: Key Image Notes (see above for its function), applied to the Image Display, the Evidence Creator and the Portable Media Creator actors. RWF: Reporting Workflow (see above for its function), applied to the Report Creator and the Report Reader actors. PDI: Portable Data for Imaging (see above for its function), applied to the Image Display, the Report Creator and the Report Reader actors, for allowing creation of CD or DVD, containing all clinical data collected during the clinical procedure to be handled to patient. To achieve the second goal, i.e., the exchange of documents and images across healthcare enterprises, two new actors: (7) Document Consumer and (8) Imaging Document Consumer have been implemented, and the following integration profiles have been associated to them: XDS: Cross-Enterprise Clinical Document Sharing, applied to the Document Consumer actor, allowing exchange of health information, with adequate access control, making it retrievable for all health professional figures, external to the health enterprise, co-operating in guaranteeing citizen s health. XDS-I: Cross-Enterprise Document Sharing for Imaging (see above for its function), applied to both the actors, i.e., Document Consumer and Imaging Document Consumer. Finally, two other new actors have been implemented in HDW2 to prepare the Patient CD (a CD containing all the relevant images, the medical report and the viewer, and substituting the films and the paper-based report) and to print images and data: (9) Portable Media Creator: to prepare contents to be written on some media storage support, and execute writing; (10) Print Composer: to drive a DICOM printer, including transmission of possible masks to be applied upon images. The Portable Data for Imaging (PDI) profile (see above for its function) has been applied to both these two actors. As for O3-DPACS, the development of most of these new profiles required a preliminary research for finding and setting up the most suitable solution and interpretation of the profiles themselves, with much cooperative work with other research centres at international level. More information about the HDW2 IHE Integration Statements and DICOM Conformance Statements can be found in the HDW2 web pages [80] and on the official IHE website [66]. 3. Results 3.1. O3-DPACS Under the shell, O3-DPACS behaves as a layered system, so taking full advantage of Java technology, by empowering Fig. 4. Layered view of O3-DPACS. the portability of Java and the scalability of J2EE application servers, which can for example rely on clusters without changing the Java application code. Fig. 4 shows its basic structure. Whatever the operating system, O3-DPACS relies on a J2EE compatible application server, which is based on the JVM (Java Virtual Machine). The Java layer allows the access to Windows or Linux file-systems. Moreover, using the JDBC interface, the system is also compatible with all the DBMSs, which provide a JDBC driver. Fig. 5 shows the deep structure of O3-DPACS. The system is packaged in a unique EAR (Enterprise Application archive), as a J2EE application. It is deployed on the application server and has all the modules inside it, except a start-up class pictured in the lowest right corner of Fig. 5. This class creates a Server for Managed Beans (MBean server) and registers all the O3-DPACS services on it. This choice allows the complete control of parameters of the classes responsible for services [81], e.g., listening ports for DICOM Services or other configuration parameters, such as communication timeouts, AE-Title filters and parameters publication on a secure web page, according to the JMX (Java Management extensions) standard specifications. Hence, operators can change the configuration from a web page without restarting the system and multiple authentication methods can be implemented to avoid misuse of this critical feature. This approach gives also some guarantees on the portability to various application servers, which usually support JMX modules as a part of J2EE specifications. The implemented servers are six: DICOM Storage, DICOM Modality Performed Procedure Step (MPPS), DICOM Query, DICOM Retrieve, DICOM Storage-Commitment, and HL7 server. These servers are responsible for receiving a message and understanding whether it is under their responsibility. If so, they call for business logic. It is important to notice that all the servers (either DICOM or HL7) implement either the normal mode or the secure mode through the TLS protocol, where certificates need to be declared. The business logic is written as an Enterprise Java Beans (EJB) layer and its main concern is parsing the message, updating the database, eventually storing data on the file-system and doing additional tasks as forwarding the message to another system. The Java Beans architecture achieves two goals [82]. It fosters scalability since objects are accessed either from local or remote

10 400 P. Inchingolo et al. / Computerized Medical Imaging and Graphics 30 (2006) Fig. 5. Deep class organization of O3-DPACS application. interfaces, which for example are a requisite to distribute the system on a cluster. Moreover, the EJB interfaces can be easily called from web, via a normal web access, as well as via other applications implementing JNDI (Java Naming and Directory Interface) or RMI (Java Remote Method Invocation); alternatively, they can be used as a web service, again either from an application or from a browser. Since there are six servers, there are also six dealers which, coupled together, serve as a two-layers structure implementing DICOM Storage Service, DICOM Query Service, DICOM Retrieve Service, DICOM Modality Performed Procedure Step Service and DICOM Storage Commitment Service plus the HL7 Service. The dealers access the database through JDBC methods; routing to a data-source is defined externally in some deployment descriptors of the application server. This makes it easy to switch between DBMSs without altering the application code. The auto-routing module is accessed from the storage server and implements Java Messaging System (JMS technology) to let the routing mechanism be asynchronous and also a totally independent module. It has proved to be a really top interest feature for information systems in transition, for the flexibility and the possibility of interaction with other archiving facilities. As already mentioned, the configuration tool (Fig. 6) interacts with the database for modifying the O3-DPACS configuration such as node and modality configurations or SOP (Service Object Pair) class adding or for auto-routing routes definition. It is packed with O3-DPACS as a Web Application archive (WAR). The administration tool, instead, has been developed to have a view on the whole database and let the remote operator find faulty records and manage the database. It avoids the installation of a database client on the system and is perfectly included as a WAR archive on the application server web container. The tools have been programmed in order to be easily internationalized by means of changing character set and translation. Several languages have been already installed. Finally, thanks to the implementation of the MBean Server, which registers all O3-DPACS modules, the tool can act on the database and force servers and dealers to read changes and keep them updated as the operator is working without any stop of the server HDW2 The implementation of HDW2 Workstation (Fig. 7) in many different environments has confirmed its power as a platform for both clinical and medical solutions. Moreover, it can fit in IHE or not-ihe hospital environments and it makes easy to add software modules which help the user in his/her job: it is possible, for example, to add, inside HDW2, several different authentication systems, or to deny some application modules. HDW2 therefore behaves as a system exhibiting high modularity, configurability and adaptability, acting as a solution provider for clinical issues, following clinical environment trends and fitting its features to fulfil these needs, for example, those related to the IHE profiles in the Information Technology domain (ITI), which become more and more important and they are felt in a good way by the community [71]. HDW2 is the first medical workstation in the world designed to manage any type of images, data and waveforms in a multiple

11 P. Inchingolo et al. / Computerized Medical Imaging and Graphics 30 (2006) Fig. 6. PACS configuration tools. medical environment: in radiology, in cardiology, in anatomopathology, in pneumology, etc. Although being so much open to any data type, it maintains full adherence to IHE profiles specification. Furthermore, the internationalization of HDW2 allows its use even in a multi-language hospital or territorial body. This is a feature particularly important in multi-lingual regions and countries, as Friuli Venezia Giulia in Italy or Switzerland, which both have four official languages: Italian, German and Ladin/Friulian, plus Slovenian in the first one and French in the second one. Fig. 7. The HDW2 workstation. HDW2 is now being equipped also with other character sets, as Cyrillic and Arabian. The integration in HDW2 of many add-on modules, as the O3-3D visualization and manipulation module, the O3-TEBAM electrical reconstruction one and the O3-MIRC connection to the Medical Imaging Resource Centres, adds much power to the workstation and large interest in medical doctors O3-DPACS & HDW2 validation O3-DPACS and HDW2, which are the current results of the studies and of the concept applications described in this paper, have now passed through a huge test phase which has validated most of the assumptions done during the project. The first set of tests for the achievement of interoperability has been done at the IHE 2005 Connectathon in Amsterdam and the second one at the IHE 2006 Connectathon in Barcelona. O3- DPACS and HDW2 gained compliance to the above mentioned profiles, summarized in Table 2, passing more than 300 tests with most of the European market brands, and ranking as a number of successfully tested profiles at the 9th position over more than 200 companies. Since Summer 2005, O3-DPACS is fully functional for production in Regione Veneto at the University Hospital and in the Healthcare Enterprise of Padova, where it is managing 9 GB of images every day, according to last statistics, including multislice images. In Regione Friuli Venezia-Giulia, at the Hospital- University Enterprise of Trieste, Cattinara Hospital, O3-DPACS and HDW2 have fully substituted in November 2005 the previ-

12 402 P. Inchingolo et al. / Computerized Medical Imaging and Graphics 30 (2006) ous installation of the DPACS server and workstations which were in service since 1997; it is also used within the PRIN Italian National Research Project TEBAM (True Electrical Brain Activity Mapping), coordinated by the Group of Bioengineering of Trieste [78,79]. In Tuscany, O3-DPACS is installed at the Hospital-University Enterprise of Pisa in the Hospital S. Chiara and in the Hospital Cisinello, where it manages more then 10 GB of images every day; furthermore, it is used in Pisa also within the TEBAM project. In 2006 another large O3-DPACS installation is programmed at the Hospital S. Carlo of Milano. The success of O3-DPACS is now spreading also across many European (as, e.g., Slovenia, in particular at the University Hospitals of Maribor and Ljubljana) and extra-european countries (as, e.g., in Iran at the Noor Medical Imaging Centre of Teheran). HDW2 is currently used, in addition to the Hospitals of Trieste, in many other Hospitals of Regione Friuli Venezia Giulia, where it is installed and managed by Insiel and in some of the O3-DPACS sites. Experimental installations in many countries are planned for the next months. Outside Radiology or other clinical production departments (as Cardiology), images, data and signals are distributed to clinicians in the hospital and to other hospitals using DICOM and HL7 protocols connecting directly to HDW2, or, alternatively, interoperating thanks to the IHE profiles with full hospital or inter-hospital Electronic Health Record (EHR) Systems. For example, in the Hospital of Padova O3-DPACS interoperates at 100% with the ITACA hospital-wide EHR system provided by Dianoema Group. The implementation, with original solutions, of the XDS (Cross-Enterprise Clinical Document Sharing) and the XDS-I (Cross-Enterprise Document Sharing for Imaging) profiles, has opened a very new scenario. Images and data can be exchanged very easily within any territorial environment Java choice and performances The choice of Java Technology has been the result of accurate evaluations, since still today there is a debate about Java slowness, especially when compared to C++ applications. The following two different aspects have been in particular evaluated, regarding the two major areas: enterprise applications design and code execution. Java Enterprise Technology allows the developer to focus on designing business logic, while its intrinsic scalability takes care about getting the most from the current hardware. As an example, the internal support for clustering allows the developer to decide, in a multiprocessor environment, which resources should be assigned to the application modules, tailoring the use of hardware resources on the application computational needs. Hence, careful design strategies can close most of the performance gap, as we actually experimented in our studies and experimental work. From user perception point of view, the HDW2 Java image manager concludes, under services condition, an exam search within the time of 1 s for a mean query of 40 results among 500,000; retrieval time for the most common exams is under 10 s. These results are aligned with different PACS implementations using non-java technologies, calculated on the same hardware. Moreover, benchmarks state that, mainly thanks to JIT (Just In Time compilers), which avoids interpretation of critical code portions, the gap between execution of Java and C++ code is now closing [83,84]. To better assess Java on graphics, we are making benchmarks on the main graphical libraries as JAVA2D and JAI, and we are looking also for different ones. The performed benchmarks have shown the Java strength on critical and stress graphics procedures that are the key for the use of Java for medical imaging Implementation, training and service model Two different models are applied to diffuse, implement and assure training and full service for O3-DPACS and HDW2. Implementations carried out by Insiel follow the line of distribution and maintenance proper of Insiel, a 650-employes company (fully owned by Regione Friuli Venezia Giulia) managing all the healthcare and other public networks and services of Regione Friuli Venezia Giulia as well as some public services in most of the other regions in Italy. These Insiel implementations and services are characterized by a tight cooperation with HECE and the Bioengineering and ICT group of the University of Trieste and, in the premises, with the Open Three Consortium, also for the update and the improvement of the R&D products and for the design of original distributed solutions. The second model applies to the R&D products under the umbrella of the Open Three Consortium Project (O3-C). The O3-C Community is made by all the institutions having an agreement with HECE: they are, in particular, those belonging to the international networks ABIC-BME (Adriatic Balcanic Ionian Cooperation in Biomedical Engineering) and ALADIN (Alpe Adria Initiative Universities Network), and the institutions about 55 healthcare and industrial enterprises or governmental agencies having a bilateral agreement active with HECE. In the O3 Community, an O3 Users Community and an O3 Developers Community are identified. Every member of the O3 Community can in principle ask to participate to both communities. The Developers community, based currently 90% on the Universities of Trieste and Padova under the responsibility and administration of HECE (but now growing up with many other contributions), provides the active members of the Users Community with all the necessary project design, site analysis, implementation, logging, authoring, bugs solving, and high-level 7/7 24/24 full service. Additionally, training is highly cared, starting with preparing clinical engineering professionals at three different levels, offering both traditional and e-learning courses with particular skills in Clinical Informatics, Health Telematics, e-health integration standards and IHE-based interoperability, and providing also specific courses and training on site. Since this service work is growing a lot, O3 Enterprise, a spin-off from the University of Trieste, is now being constituted, to offer services of implementation, management, customization and integration to the healthcare enterprises in Italy and abroad. Other spin-off companies with Universities abroad (e.g. with the University of Maribor in Slovenia) are also under study.

13 P. Inchingolo et al. / Computerized Medical Imaging and Graphics 30 (2006) Discussion and conclusions The experience of 15-years study of the Bioengineering and ICT group of Trieste on the health systems integration by means of ICT technologies, from the hospital department to the networked reality of the single citizen in the e-health context of the future information-based society, has shown, thanks to the practical experimentation of the above described solutions, that some methodological and organizational key elements are really strategic for the success of the e-health integration process. One of these key-elements is surely interoperability, which nowadays is largely promoted by the IHE world-wide project. Another key-element, especially for a world-wide integration including industrial and transitional countries, is the availability of open-source products, accompanied by a serious program of maintenance and of continuous software upgrade. The availability of flexible images and data displays, able to interoperate and work with any type of medical data by using different types of terminals, from MAC or Windows workstations to the new generation cellular phones, has been proved to be another key aspect. With regard to this last issue, the code portability for client and server software has been assured in our project by the definitive choice of Java technology. From the point of view of the organization of our cooperative work with other user and developer centers, the initiative of the Open Three (O3) Consortium [84] has been proved in the last months its real efficiency and efficacy. All O3 sub-systems are harmonically developed and can be scaled at any range, up to national and international dimensions. O3 is developed completely as Open Source and with Java and Web technologies to facilitate its re-use and portability, fostering a wide diffusion in Italy and abroad. It is data-base, OS, HW and language independent, and 100% compliant with the world-wide interoperability initiative IHE. The choice of Open Source as the leading solution for the future of e-health anticipates a common trend in the industrialized and political world, recently evidenced by the position assumed by the Department of Health & Human Services and the Department of Defence of Unites States at the Open Source Strategy for Multi-Centre Image Management Workshop, held in 6 9 March 2006 at Las Vegas (USA), by the decision announced by the world s biggest industries at the OSDL Joint Initiatives Face to Face Meeting Review Health Care Information Exchange, held in May 2006 at Sophia- Antipolis (France), and finally by the European Union with the Riga Declaration signed during the Intergovernmental Meeting of the European Commission ICT for an Inclusive Society, held in June 2006 at Riga (Latvia). It is interesting to underline that O3-DPACS and O3 Consortium have been presented by invitation at all these three events. Another important point is that O3 s bricks are built according to IHE Actors. O3 s information flows are totally compliant with IHE Integration Profiles. The success of HDW2 as a universal workstation led also to the development of a simple but efficient and revolutionary open-source radiology workstation, called O3-RWS that will be fully available before the end of In conclusion, the DPACS project, in particular the new O3- DPACS Image-Data Manager/Archiver and the HDW2 Image- Data Display, and now the O3 Consortium, seem to represent a significant contribution to really push the e-health integration in Europe and the world. The results obtained in promoting the e-inclusion in Europe in relation to the health, also in complementary fields, in particular in multilingual and transnational e-learning in health, as the ISCELS e-learning platform [85] awarded in May 2006 by the Italian Inter-ministerial Committee for e-inclusion and accessibility corroborate the European value of the DPACS project. Acknowledgements Work supported by the Higher Education in Clinical Engineering (HECE), University of Trieste, by the Italian Ministry of University and Research, National project PRIN 2004 n TEBAM, by the Central European Initiative, by Insiel Spa, by the Interuniversity Consortium CINECA, Casalecchio di Reno (Bologna), by the European Community and Regione Friuli Venezia Giulia, INTERREG IIIA Italy-Slovenia, Axis 3, Measure 3.2, Action 3.2.4, Grant # BAFVG ISCELS. The Authors want to thank O3-PD, the Section of Padova of the O3 Consortium (headed by Prof. Giampietro Feltrin and Dr. Claudio Saccavini), for having developed a powerful backup module for O3-DPACS and for having deeply tested and validated O3-DPACS in the Hospitals of Padova; thanks also to the Departments of Radiology of Pisa (resp. Prof. Davide Caramella) and of Trieste (resp. Prof. Maria Assunta Cova) for having tested and validated O3-DPACS, and to Insiel spa (resp. Dr. Mauro Masè) for having tested and validated HDW2 in the hospitals of Regione Friuli Venezia Giulia. References [1] Cao X, Huang HK. Current status and future advances of digital radiography and PACS key components of system integration in modern healthcare. IEEE Eng Med Biol Mag 2000;5:80 8. [2] Noumeir R. DICOM structured report document type definition. IEEE Trans Inf Tech Biomed 2003;7(4): [3] Pavlopoulos SA, Delopoulos AN. Designing and implementing the transition to a fully digital hospital. IEEE Trans Inf Tech Biomed 1999;3(1):6 19. [4] Hacklander T, Kleber K, Martin J, Mertens H. DICOM router: an open source toolbox for communication and correction of DICOM objects. Acad Radiol 2005;12(3): [5] Arenson RL, Andriole KP, Avrin DE, Gould RG. Computers in imaging and health care: now and in the future. J Digit Imaging 2000;13(4): [6] Gibaud B, Garfagni H, Aubry F, Pokropek AT, Chameroy V, Bizais Y, et al. Standardization in the field of medical image management: The contribution of the MIMOSA model. IEEE Trans Med Imaging 1998;17(1): [7] Freiherr G. PACS: A premature paradise. (Pictorial Archiving and Communications Systems). Comput Graph World 1990;13(4):52. [8] Inamura K, Kousaka S, Yamamoto Y, Sukenobu Y, Okura Y, Matsumura Y, et al. PACS development in Asia. Comp Med Imaging Graph 2003;27:121 8.

14 404 P. Inchingolo et al. / Computerized Medical Imaging and Graphics 30 (2006) [9] Inamura K, Konishi J, Nishitani H, Kousaka S, Matsumura Y, Takeda H, et al. Status of PACS and technology assessment in Japan. Comp Meth Prog Biomed 2001;66:5 15. [10] Siau K. Health care informatics. IEEE Trans Inf Tech Biomed 2003;7(1):1 7. [11] Wong STC, Huang HK. A hospital integrated framework for multimodality image based management. IEEE Trans Sys, Man, Cybern (A) 1996;4: [12] Muller H, Michoux N, Bandon D, Geissbuhler A. A review of contentbased image retrieval systems in medical applications-clinical benefits and future directions. Int J Med Inform 2003;7(1):1 7. [13] Kondoh H, Washiashi C, Sasagaki M, Arisawa J, Nakamura H, Inamura K. Development and evaluation of PC-based HIS-RIS-modality-PACS coupling: the results of evaluation of initial stage with personal computer application. Comp Meth Prog Biomed 1998;57:63 8. [14] Inamura K, Umeda T, Sukenobu Y, Matsuki T, Kondo H, Takeda H, et al. HIS:RIS contribution to image diagnosis and maximization of efficacy of PACS when coupled with HIS:RIS. Comp Meth Prog Biomed 1998;57:41 9. [15] Inchingolo P. PACS in today s health care. Invited paper. In: Lloyd N, Fehimovic M, Gahan E, Cooper E, editors. Next-generation Health Care: London, pp [16] Ratib O, Hoehn H, Girard C. Parisot C. PAPYRUS 3.0: DICOM-compatible file format. Med Inf 1994;19(2): [17] Wong STC, Tjandra D, Wang H, Shen W. Workflow-enabled distributed component-based information architecture for digital medical imaging enterprises. IEEE Trans Inf Tech Biomed 2003;7(3): [18] Chang Z, Mei S, Gu Z, Gu J, Xia L, Liang S, et al. Realization of integration and working procedure on digital hospital information system. Comp Stand Int 2003;25: [19] Bui AAT, Dionisio JDN, Morioka CA, Sinha U, Taira RK, Kangarloo H. DataServer: an infrastructure to support evidence-based radiology. Acad Radiol 2002;9: [20] Inamura K, Kondoh H, Takeda H. Development and operation of PACS/Teleradiology in Japan. IEEE Commun Mag 1996;4: [21] Inchingolo P, Belloni R, Bergamasco S, Bosazzi P, Mininel S, Miniussi E, et al. New trends of the DPACS project. In: Niinimaki J, Ilkko E, Reponen J, editors. Proceedings of the 20th EuroPACS Annual Meeting. Oulu University Press; p [22] Inchingolo P, Pozzi Mucelli R, editors. EuroPACS-MIR 2004 in the Enlarged Europe. EUT, Trieste, ISBN: Also available on line at [23] Bourka A, Polemi D, Koutsouris D. Interoperability among healthcare organizations acting as certification authorities. IEEE Trans Inf Tech Biomed 2003;7(4). pp. 364, 377. [24] alsafadi Y, Lord WP, Mankovich NJ. PACS/information systems interoperability using enterprise communication framework. IEEE Trans Inf Tech Biomed 1998;2(2):42 7. [25] Chronaki CE, Chiarugi F. Interoperability as a quality label for portable and wearable health monitoring systems. Stud Health Technol Inform 2005;117: [26] Drury P, Dahlman B. Open source approaches to health information systems in Kenya. World Hosp Health Serv 2005;41(3):36 9. [27] Rosset A, Spadola L, Pysher L, Ratib O. Informatics in radiology (inforad): navigating the fifth dimension: innovative interface for multidimensional multimodality image navigation. Radiographics 2006;26(1): [28] Hacklander T, Martin J, Kleber K. Informatics in radiology (inforad): an open source framework for modification and communication of DICOM objects. Radiographics 2005;25(6): [29] Lindahl F, Torgersson O. mgen an open source framework for generating clinical documents. Stud Health Technol Inform 2005;116: [30] Kalra D, Beale T, Heard S. The openehr Foundation. Stud Health Technol Inform 2005;115: [31] Saranummi N. PICNIC architecture. Stud Health Technol Inform 2005;115: [32] Oyri K, Murray PJ. osni.info using free/libre/open source software to build a virtual international community for open source nursing informatics. Int J Med Inform 2005;74(11/12): [33] Rosset A, Spadola L, Ratib O. OsiriX: an open-source software for navigating in multidimensional DICOM images. J Digit Imaging 2004;17(3): [34] Paindaveine IY. Open source in the European union: a future way for health applications? Stud Health Technol Inform 2003;96: [35] McDonald CJ, Schadow G, Barnes M, Dexter P, Overhage JM, Mamlin B, et al. Open source software in medical informatics why, how and what. Int J Med Inform 2003;69(2/3): [36] Ratib O, Swiernik M, McCoy JM. From PACS to integrated EMR. Comput Med Imaging Graph 2003;27(2/3): [37] Langer S. OpenRIMS: an open architecture radiology informatics management system. J Digit Imaging 2002;15(2):91 7. [38] Yackel TR. How the open-source development model can improve medical software. Medinformation 2001;10(1): [39] Inchingolo P, Bosazzi P, Cicuta D, Faustini G, Barbaro A, Miniussi E, et al. DPACS-2004 becomes a Java-based Open-Source Modular System. In: Inchingolo P, Pozzi Mucelli R, editors. EuroPACS-MIR 2004 in the Enlarged Europe p ISBN: [40] Miniussi E, Bosazzi P, Inchingolo P, Masè M, Vittor A. HDW2: a powerful and customization-friendly java-based Dicom Workstation. In: Inchingolo P, Pozzi Mucelli R, editors. EuroPACS-MIR 2004 in the Enlarged Europe p ISBN: [41] Saccavini C. The MARIS project: open-source approach to IHE radiological workflow software. In: Inchingolo P, Pozzi Mucelli R, editors. EuroPACS-MIR 2004 in the Enlarged Europe p ISBN: [42] Inchingolo P, Saccavini C, Beltrame M, Bosazzi P, Faustini G, Poli A, et al. The Open Three (O3) Consortium Project of the Universities of Trieste and Padova. In: Proceedings of the 19th IEEE MIPRO Convention, vol p [43] Giribona P, Bravar D, Stacul F, Ukovich W. PACS experiences in Trieste. Int J Biomed Comput 1992;30(3 4): [44] Diminich M, Inchingolo P, Magliacca F, Martinolli N. Versatile and open tools for LAN, MAN and WAN communications with PACS. In: Held KD, Brebbia CA, Ciskowski RD, Power H, editors. Computational biomedicine. Southampton: Computational Mechanics Publications; p [45] Inchingolo P. From the GNBTS-NET to the biomedical communication network of Trieste. In: Bracale M, Denoth F, editors. Proceedings of the MEDICON 92. Pisa: Area di Ricerca, CNR; p [46] Diminich M, Inchingolo P, Martinolli N, Dalla Palma L, Giribona P, Pozzi R, et al. ARIS: an autonomous remote image station offering both local and remote IMACS/RIS resources in an open environment. In: Proceedings of the Third European Conference on Engineering Medicine p [47] Fioravanti F, Inchingolo P, Valenzin G, Dalla Palma L, Vallon R. The DPACS Project at the University of Trieste: architecture and working environment. In: Proceedeedings of the 14th International EuroPACS Meeeting [48] Fioravanti F, Inchingolo P, Valenzin G, Dalla Palma L, Vallon R. HIS-RIS- PACS integration with the DPACS Project at the University of Trieste. In: Proceedings of the 14th International EuroPACS Meeeting [49] Fioravanti F, Inchingolo P, Valenzin G, Dalla PL. The DPACS Project at the University of Trieste. Med Informat 1997;22(4): [50] Inchingolo P. Trieste and Region Friuli-Venezia Giulia: an example of fullymanaged health telematics system the new Telematic Network of 2000 s Trieste. IFMBE Proceedings 2001;1: [51] Hussein R, Engelmann U, Schroeter A, Meinzer HP. Internationalization of healthcare applications: a generic approach for PACS workstations. Meth Inf Med 2004;43(2): [52] Tahayori B, Soltanian-Zadeh H, Zoroofi RA. A PC-based PACS display workstation with a language transparent interface. J Digit Imaging 2002;15(Suppl 1): [53] Raman B, Raman R, Raman L, Beaulieu CF. Radiology on handheld devices: image display, manipulation, and PACS integration issues. Radiographics 2004;24(1):

15 P. Inchingolo et al. / Computerized Medical Imaging and Graphics 30 (2006) [54] Kim Y, Fahy JB, DeSoto LA, Haynor DR, Loop JW. Development of a PC-based Radiological Imaging Workstation. J Digit Imaging 2003;16(1): [55] Doyle AJ, Gunn ML, Gamble GD, Zhang M. Personal computer-based PACS display system: comparison with a dedicated PACS workstation for review of computed radiographic images in rheumatoid arthritis. Acad Radiol 2002;9(6): [56] Gillessen C, Ricke J, Haderer A, Zielinski C, Teichgraber U, Ehrenstein T, et al. A windows based viewing application for image postprocessing in DICOM format. Stud Health Technol Inform 2000;77: [57] Ernst R, Le VT, Kawashima A, Caskey C, Zelitt D, Tamm E, et al. A picture archiving and communications system featuring multiple monitors using Windows98. J Digit Imaging 1999;12(2 Suppl 1): [58] Mehta A, Dreyer K, Thrall J. Enhancing availability of the electronic image record for patients and caregivers during follow-up care. J Digit Imaging 1999;12(2 Suppl 1): [59] Ghosh S, Andriole KP, Avrin DE, Arenson RL. Optimization of a low-cost truly preemptive multitasking PC diagnostic workstation. J Digit Imaging 1997;10(3 Suppl 1): [60] Cosic D. An open medical imaging workstation architecture for platformindependent 3D medical image processing and visualization. IEEE Trans Inf Technol Biomed 1997;1(4): [61] Wang M, Lau C, Matsen III FA, Kim Y. Personal health information management system and its application in referral management. IEEE Trans Inf Tech Biomed 2004;8(3): [62] Arenson RL, Chakraborty DP, Seshadri SB, Kundel HL. The digital imaging workstation. J Digit Imaging 2003;16(1): [63] Slomka PJ, Elliott E, Driedger AA. Java-based remote viewing and processing of nuclear medicine images: toward the imaging department without walls. J Nucl Med 2000;41(1): [64] Foord KD. PACS workstation respecification: display, data flow, system integration, and environmental issues, derived from analysis of the Conquest Hospital pre-dicom PACS experience. Eur Radiol 1999;9(6): [65] Tamm EP, Raval B, West OC, Dinwiddie S, Holmes R. Evaluating the impact of workstation usage on radiology report times in the initial 6 months following installation. J Digit Imaging 1999;12(2 Suppl 1): [66] IHE, Integrating the Health Care Enterprise. [67] Tellis WM, Andriole KP. Integrating multiple clinical information systems using the Java message service framework to enable the delivery of urgent exam results at the point of care. J Digit Imaging 2005;18(4): [68] Tellis WM, Andriole KP. Integrating multiple clinical information systems using the Java Message Service framework. J Digit Imaging 2004;17(2):80 6. [69] JBOSS, The professional Open Source Company, [70] Bandon D, Trolliard P, Garcia A, Lovis C, Geissbu hler A, Valle JP. Building an enterprise-wide PACS for all diagnostic images. Int Congress Ser 2004;1268: [71] HL-7, Health Level 7. [72] O3-DPACS, Open Three Consortium, Section of Trieste, University of Trieste. [73] Open Three (O3) Consortium, [74] The Apache Tomcat servlet container, [75] Melzer K, Kroll M, Mahnke R, Lipinski HG. Mobile access to medical 2D/3D-images and biological waveform data on PDAs and smartphones. In: Käfer J, Zündt M, editors. Proceedings of the Eighth International Workshop on Mobile Multimedia Communications Center for Digital Technology and Management TU-LMU Munich p [76] Vatta F, Bruno P, Inchingolo P. Improving lesion conductivity estimate by means of EEG source localization sensitivity to model parameter. J Clin Neurophysiol 2002;19(1):1 15. [77] Mininel S, Bruno P, Vatta F, Inchingolo P. 3D functional and anatomical data visualization on auto-stereoscopic display. In: Inchingolo P, Pozzi Mucelli R, editors. EuroPACS-MIR 2004 in the Enlarged Europe p [78] Vatta F, Bruno P, Inchingolo P. Multi-region bicentric-spheres models of the head for the simulation of bioelectric phenomena. IEEE Trans Biomed Eng 2005;52(3): [79] Bruno P, Mininel S, Inchingolo P, Vatta F. An ICT-based innovative framework for 3D true electrical brain activity mapping: methodological and computational issues. In: Proceedings of the 19th IEEE MIPRO Convention, vol p [80] HDW2, HTL Dicom Workstation project, HTL, DEEI, University of Trieste. [81] Tony GT. JMX boosts J2EE Application Management Auerbach Publications. CRC Press LLC; [82] Li M, Wilson D, Wong M, Xthona A. The evolution of display technologies in PACS applications. Comput Med Imaging Graph 2003;27(2/3): [83] Lewis JP, Neumann U. Performance of Java versus C++. Computer Graphics and Immersive Technology Lab, University of Southern California. January 2000, upd zilla. [84] Mangione C. Just In Time for Java versus C++. NC World. 1 February [85] ISCELS, Interreg IIIA Italian Slovenian Cooperative E-learning Space. Paolo Inchingolo is full professor of Bioengineering, Biological Control Systems and Health Telematics at the Faculty of Engineering of the University of Trieste, Italy. He is the head of the bioengineering and ICT Group at the Department of Electrotechnics, Electronics and Informatics (DEEI) of the University of Trieste. He has a 33 years career started as assistant professor of human physiology at the Universities of Trieste and Ferrara. He is visiting professor of bioengineering at the National Institutes of Health, Bethesda, USA, since He promoted many research projects and centres in the fields of neuromodelling, clinical engineering, telemedicine, medical informatics and advanced biomedical instrumentation. In 1991, he founded the international program Higher Education in Clinical Engineering (HECE) of which he is the director, signing more than 60 agreements with research centres, hospitals, companies and foreign universities. He is founder and coordinator of the University Adriatic- Balcanic-Ionian Cooperation on Biomedical Engineering (ABIC-BME) and Delegate of the Alpe-Adria Universities Network (ALADIN). He is Chair of the IHE Transnational Committee for Central & Eastern Europe and President of the Open Three Consortium. He is author of 360 papers and member of IEEE, AEI, AIIMB, IFMBE, EUROPACS societies. His web portal: Marco Beltrame received the Laurea degree in electronic engineering (biomedical curriculum) in 2004 at the University of Trieste, Italy. He is working toward the PhD degree in information engineering Bioengineering curriculum at the same university. Since 2005 he is working with the Bioengineering and ICT group at the Department of Electrotechnics, Electronics and Informatics (DEEI) of the University of Trieste, Italy, within the Health Telematics Lab (HTL) and the Open Source Lab (OSL). Since July 2006, he is a postdoctoral fellow at DEEI. He is referent for O3-DPACS in the frame of the O3-Consortium project. He is involved in research activities in the fields of e-health, in particular in applications of telemedicine, clinical informative systems, sharing of clinical data and images. Pierpaolo Bosazzi received in 1990 the computer science diploma and in 1997 the BSc in informatics and automation engineering at the University of Trieste, Italy. He has been employed as system and network administrator from 1990 to 1995 at the Centre for Research and Studies on Biomedical and Health Technologies (CRSTBS), AREA Science Park, Trieste, Italy, and at IRCCS Burlo Garofolo Hospital in Since 1997 he is employed as permanent technician at the Clinical Unit of Radiology, University of Trieste, Italy, as administrator of the DPACS/O3-DPACS system. Davide Cicuta was born in Trieste, Italy, in In 2005, he received the Laurea degree in electronic engineering at the University of Trieste. Since 2003 he worked for the Bioengineering and ICT group at the Department of Electrotechnics, Electronics and Informatics (DEEI) of the University of Trieste, Italy, within the Health Telematics Laboratory (HTL). Giorgio Faustini was born in 1982 in Trieste, Italy. He graduated in information engineering in 2004 at the University of Trieste, Italy. He is working toward his Mag. Laurea graduation in Clinical Engineering on March Since September 2003 he is a fellow of the Health Telematics Laboratory (HTL) and of the Open Source Lab (OSL) of the Bioengineering & ICT Group at the Depart-

16 406 P. Inchingolo et al. / Computerized Medical Imaging and Graphics 30 (2006) ment of Electrotechnics, Electronics and Informatics (DEEI) of the University of Trieste, Italy. His main research interests are in the field of medical imaging, complex software architecture, 3D medical reconstruction and distributed computing. Stefano Mininel is a postdoctoral fellow of the bioengineering and ICT group at the Department of Electrotechnics, Electronics and Informatics (DEEI) of the University of Trieste, Italy, where he works since 2001 within the Neuromodelling Lab (NML) and the Open Source Lab (OSL). He received his Laurea degree in electronic engineering (biomedical curriculum) from the University of Trieste in 2001 and the PhD degree in information engineering Bioengineering curriculum in His research interests are: protocols in medicine, modelling of bioelectrical activity, parallel computing, 3D and stereographic visualization. Andrea Poli is a postdoctoral fellow of the Bioengineering and ICT group at the Department of Electrotechnics, Electronics and Informatics (DEEI) of the University of Trieste, Italy, where he works since 2003 within the Health Telematics Lab (HTL) and the Open Source Lab (OSL). He received the Laurea degree in electronic engineering (biomedical curriculum) from University of Trieste in His research interests are: protocols in medicine, java programming, development of IHE compliant systems, e-learning. Federica Vatta is assistant professor of the bioengineering and ICT group at the Department of Electrotechnics, Electronics and Informatics (DEEI) of the University of Trieste, Italy, since July She received the Laurea degree (cum laude) in electronic engineering and the PhD degree in information engineering bioengineering curriculum from the University of Trieste, Italy, in 1998 and 2002, respectively. She has been awarded by 3 international awards for best presentation of her scientific results and one national prize for her PhD thesis from the National Bioengineering Group (GNB). Since January 2001 to July 2005 she has been a postdoctoral fellow at DEEI of the University of Trieste. She is currently the didactic manager of the Higher Education in Clinical Engineering international program at the Faculty of Engineering of the University of Trieste. Her main research interests are medical imaging, 3D medical reconstruction, distributed computing and electromagnetic brain activity mapping.