A Data Grid Model for Combining Teleradiology and PACS Operations

Transcription

1 MEDICAL IMAGING TECHNOLOGY Vol.25 No.1 January 特 集 論 文 / 遠 隔 医 療 と 画 像 通 信 A Data Grid Model for Combining Teleradiology and Operations H.K. HUANG *, Brent J. LIU *, Zheng ZHOU *, Jorge DOCUMET * Abstract The use of teleradiology for X-ray image communication and display was introduced as early as and the concept was conceived in , 3 ; after more than twenty years of technological advancement and refinement, both teleradiology and have become indispensable components in today s healthcare delivery system. Although both teleradiology and share many common technological components and workflow profiles, these two imaging-based systems are operated independently and are not readily integrated. This paper compares the commonalities and differences between these two systems, and suggests a method for their integration in terms of image archival, reporting, and workflow using a Data Grid model. Key words:, Teleradiology, Data Grid,, IHE Med Imag Tech 25(1): 7-12, Introduction Teleradiology is a subset of telemedicine operations focusing on remote diagnosis of medical images. Teleradiology utilizes computer, display and telecommunication technologies for radiologists at workstations (s) to make diagnosis from clinical images generated from remote examination sites. The diagnostic report is sent back to the examination site where a primary physician can provide proper treatments to the patient immediately. Teleradiology operation can be very simple or extremely complex as shown in Table 1. Three factors contribute to the complexity: 1) Historical images are required for comparison; 2) Information from the radiology information system (RIS) is needed; and 3) archive is necessary. Teleradiology is relatively simple to operate when neither image and information retrieval, nor archival of images and reports are required. However, when both archival and retrieval are required, the operation becomes extremely complex. In this paper, we try to address the issue How to combine images and reports from teleradiology with those from an outside archive, like a? as a means to systematically analyze the most complex teleradiology operation model shown in Table 1. We first define the pure teleradiology model, then the combined teleradiology and model. The former model allows us to appreciate the workflow and workload of multiple sites teleradiology operation. The latter model addresses the global issue related to the earlier question. Table 1 Four modes of teleradiology operation according to their complexity. Historical s/ris Archive Most simplistic no no Simplistic yes no Complex no yes Most complex yes yes * Processing and Informatics Laboratory, Department of Radiology, Keck School of Medicine, University of Southern California 4676 Admiralty Way, Suite 601, Marina del Rey, California hkhuang@aol.com receive: August 17, accept: December 4, 2006.

2 8 Med Imag Tech Vol.25 No.1 January Materials and methods 1)The pure teleradiology model with multiple sites Teleradiology can function as a pure teleradiology operation shown in Fig. 1. In this operation, the teleradiology management center serves as the operations manager. It receives images from different imaging centers, 1,, N, keeps a record but not the images, routes images to radiologists at different expert centers, 1,, M for reading based on the Supply and Demand criterion. s come back to the management center, the transactions are recorded, and the reports are forwarded to appropriate imaging centers. The management center is also responsible for the billing and other administrative functions like image distribution and workload balancing. 2) and teleradiology combined model Teleradiology applications and can be combined together as a unified healthcare enterprise operation with workflow shown in Fig Two major components in the combined model are the (upper rectangle), and the pure teleradiology model (lower rectangle). A. Pure Teleradiology Model: center sends images to the expert center for reading as in the pure teleradiology model shown in Fig. 1. The workflow step is depicted as No. 7 in Fig. 2. B. Radiologists assist Teleradiology Reading: Outside imaging centers (1) send exams to s (2) for Radiologists to read. s are sent to the database gateway for record (3), and/or to the expert center (4). C. radiologists assist reading: sends exams to expert center radiologists to read (5). The expert center returns report to the database gateway (6). The combined Teleradiology and model is mostly used in an enterprise level Healthcare system with satellite imaging centers, or in back-up radiology coverage for a hospital by imaging centers. In this model, and each imaging center keep its own images; but the reports can be shared. This model addresses the sharing and archival of reports, but not image archival. 3)A Data Grid for medical imaging applications Data Grid is a service of the Grid Computing technology 4. A Data Grid 5-7 specifically designed for clinical image backup and disaster recovery was developed at IPI ( Processing & Informatics Laboratory) using the Globus Toolkit 4 (GT4) 8.This Data Grid shown in Fig. 3 was designed to utilize the strengths of grid technology along with (Picture Archiving and Communication Systems)/ (Digital and Communi- /Data Site 1 RIS Im age/ Data Site N RIS Im aging Tele-Radiology Management / Data / Data Site 1 Site M Fig. 1 A Pure Teleradiology Model with N s producing images and M s with radiologists for reading. The Management administrates the data flow and operation.

3 Med Imag Tech Vol.25 No.1 January HIS Generic Components Generic & Data Flow Gateway s 3 Modalities Acquisition Gateway CONTROLLER & Archive Server 2 Workstations 6 Application Servers 1 5 image 4 RIS with management Teleradiology Fig. 2 (top) and Teleradiology (bottom) Combined Model: Radiologists at and at imaging centers can support each other 3. Server Server SJHC SAN Data Grid HCCII SAN P1 P2 P2 P1 SJHC FT Backup Archive HCCII Simulator IPI Fig. 3 Configuration of three sites data storage systems (shaded boxes) that comprise of image data storage resources in the Data Grid. The SAN (Storage Area Network) partition P1 at each site is used for its own images, whereas P2 becomes a shared resource of the Data Grid. Workstations outside of the Data Grid can access the grid for particular services, like image query/retrieve. SJHC: Saint Johns Healthcare, HCCII: Healthcare Consultation II, University of Southern California (USC), IPI: Processing and Informatics Laboratory, USC 6. cations in Medicine) technology for storing and distributing clinical images. In particular, some / resources are embedded within the GT4 five layer grid architecture (Fig. 4) developed by IPI. These include Storage service, Query service, and Retrieve service, which are integrated with the standard protocols in addi-

4 10 Med Imag Tech Vol.25 No.1 January 2007 tion to the use of other Data Grid s. We use this Data Grid model to discuss the workflow of a combined and Teleradiology Operations at the enterprise level. 3. Results 1) and teleradiology in an enterprise level operation using the Data Grid model and teleradiology combined model described in Fig. 2 can be extended to the enterprise level PAC systems and teleradiology operation. This model becomes popular in today s enterprise healthcare delivery system. Data Grid concept can be used to consolidate images/data obtained from both PAC systems and teleradiology operations into one unified image/report archive system for storage and distribution shown in Fig We use the Data Grid Model to address the global question raised earlier: How to combine images and reports from teleradiology with those from outside archive? Several PAC systems would be treated in this context as images and reports from outside archive. Each in the model has a shared storage contributed to the Data Grid as described in Fig. 3. In addition to P2 partition of the SAN storage in each (See Fig. 3), the Data Grid has a major software module, Teleradiology, which directs the image and report workflow of teleradiology. Refer to Fig. 5, Telerardiology has three components: the Manager, Tele Rad Arch, and Tele Rad Backup. One function of the Manager is to determine if images from an need to be archived in the Tele Rad Arch storage. If this is the case, a backup copy would also been sent to the Data Grid, which would then create two backup copies strategically saved in the Grid (Fig. 3). The Tele Rad Backup architecture and functions are the same as one of the Back-up storages (See P2 in Fig. 3) which becomes a share resource of the Data Grid. 2)Workflow in the enterprise and teleradiology operation The workflow of images, reports, and radiologist s readings are similar to that described in Fig. 2 except Item B in Section 2.2 where images from each to be archived would be determined by the Teleradiology Manager. A copy would also be sent to the Data Grid which creates two backup copies in the Grid (Fig. 5). Two new software packages would be necessary: 1) The Teleradiology to handle the image archive as well as its backup in the Data Grid, and 2) The Application Layer (Top Layer in Fig. 4) for and Teleradiology operation which would be different from that listed in Fig. 4. Data Grid Applications Storage Query Retrieve User-Level Middleware(Collective) Index MetaData Catalog Core Middleware/Globus Toolkit 3.0 (Connectivity and Resource) Replica Location Reliable File Transfer Access Interface Security GridFTP Fabric Network I2, T1 Simulator FT Archive HCCII SAN SJHC SAN Replica MetaData Fig. 4 Five-layer architecture (Fabric, connectivity, resource, collective, and application) of the Globus Toolkit 4, and contents of the Data Grid, shaded areas are integrated with services developed at IPI. Application Layer software is designed for clinical image data recovery. See also Fig. 3 caption 6.

5 Med Imag Tech Vol.25 No.1 January Operation 1 Server Storage 1 Local Archive Teleradiology Operation 1 N N Back- Up & Teleradiology in an Enterprise Operation Using Data Grid -based Data Grid Tele Rad Backup Tele Rad Arch... IHE Teleradiogy Manager... Server Storage 1 Back- Up N Local Archive M Fig. 5 Enterprise level large-scale and teleradiology combined model using the Grid Computing technology. The center of the figure is the grid computing infrastructure consisting of the Data Grid with the standard and IHE workflow profiles. The Data Grid handles the image and reports storage, and other grid computing resources take care of the workflow and management. The top row is the enterprise consists of 1,, N PAC systems. The bottom row depicts the pure teleradiology model with 1,, M operations. Each in the model has a shared storage contributed to the Data Grid as described in Fig. 3. The Data Grid has a major component, Teleradiology, to direct the image and report workflow. Shaded areas represent the data storage contribution of PAC systems and Teleradiology to the Data Grid Conclusions Following the combined and teleradiology model, we design a Data Grid model to handle the integration of the image archive and reports in an enterprise operation. Although the Data Grid model is defined, due to the complexity of its implementation, development of using the Grid Computing technology for Enterprise level and teleradiology combined model is still in its infancy. Acknowledgement Research has been partially supported by the National Institutes of Health, USA: NIH R01 EB 00298, NIH R01 LM 07606, T32 EB00438; and MI 2. References 1 Steckel RJ: Daily X-Ray Rounds in a Large Teaching Hospital Using High-Resolution Closed Circuit Television. Radiology 105: , Duerinckx A Ed: Picture Archiving and Communication systems () for Medical Applications. First International Conference and Workshop, Proc SPIE 318. Part 1 and 2, Huang HK: and Informatics: Basic Principles and Applications. Wiley & Sons, NJ, Chervenak A, Foster I, Kesselman C et al: The Data Grid: Towards an Architecture for the Distributed Management and Analysis of Large Scientific Datasets. Journal of Network and Computer Applications 23: , Liu BJ, Zhou MZ, Documet J: Utilizing data grid architecture for the backup and recovery of clinical image data. Computerized Medical and Graphics 29: , 2005

6 12 Med Imag Tech Vol.25 No.1 January Huang HK, Zhang A, Liu B et al: Data Grid for Large-Scale Medical Archive and Analysis. Proc. the 13th ACM International Conference on Multimedia 2005, pp Liu BJ, Cao F, Zhou MZ et al: Trends in Storage and Archive. Computerized Medical and Graphics 27: , Foster I: Globus Toolkit Version 4: Software for -Oriented Systems. IFIP International Conference on Network and Parallel Computing, Springer-Verlag LNCS 3779, 2005, pp Huang HK: Teleradiology, in Webster J.G., 2 nd Ed., Encyclopedia of Medical Devices and Instrumentation, Wiley & Sons, NJ 6: , 2006 H.K. (Bernie) Huang, D.Sc., FRCR(Hon.) Professor of Radiology and Biomedical Engineering; Director, Division of Informatics, Department of Radiology; and Director MS Program, Medical and Informatics, Department of Biomedical Engineering, University of Southern California, USA; Chair Professor of Medical Informatics, The Hong Kong Polytechnic University; and Honorary Professor, Shanghai Institute of Technical Physics, The Chinese Academy of Sciences. Dr. Huang pioneered in picture archiving and communication system () research. He developed the at UCLA in 1991, and the hospital-integrated at UCSF in Dr. Huang has taught at Georgetown University, University of Iowa, UCLA, UC Berkeley and UC San Francisco, the Hong Kong Polytechnic University, and the University of Southern California. Dr. Huang was inducted into the Royal College of Radiologists, London as an Honorary Fellow, for his contribu tion in research and development in November, 1992; the American Institute of Medical and Biological Engineering as a Founding Fellow, for his contribution in medical imaging in March 1993; the Euro Society as an Honorary Member for his contribution in in October, 1996; and was elected as the Honorary President, 2003 International CARS Congress, London. Dr. Huang has been Visiting Professor in many leading universities around the world, and a Board Member in leading medical imaging manufacturers. * * *