Medical Image Archive Retrieval and Communication

Albert Wong 1 Introduction 771

S. L. Lou 2 Medical Image Information Model 772

Umscisity of Cclifornic 3 Medical Image Archive System 772

3.1 Archive Server • 3.2 Database Management Subsystem • 3.3 Storage Subsystem •

3.4 Communication Network

4 DICOM Image Communication Standard 773

4.1 Information Objects • 4.2 Service Classes • 4.3 Example of C-STORE DIMSE Service

5 Archive Software Components 774

5.1 Image Receiving • 5.2 Image Routing • 5.3 Image Stacking • 5.4 Image Archiving •

5.5 Database Updating • 5.6 Image Retrieving

6 HIS/RIS Interfacing and Image Prefetching 777

6.1 Health Level Seven (HL7) Communication Standard • 6.2 Prefetch Mechanism

7 DICOM Image Archive Standard 778

7.1 Image File Format • 7.2 Query/Retrieve Service Class Operation • 7.3 Q/R Service Class Support Levels

8 PACS Research Applications 780

9 Summary 781

References 781

1 Introduction

Archiving medical images for future retrieval allows access to patients' historical images from their previous examinations. These images can be used in clinical review and diagnosis to compare with patients' current examinations, or as a resource in medical imaging related research. This chapter describes the storage and management of medical images using current digital archiving technology, and the hardware and software requirements for the implementation of a medical image archive system to facilitate timely access to medical images in an integrated hospital environment.

Medical images are produced by a wide variety of imaging equipment, such as computed tomography (CT), magnetic resonance imaging (MRI), ultrasound (US), computed radiography (CR), nuclear medicine (NM), digital subtraction

Portions of this chapter reprinted from Wong A, Huang HK, Arenson RL, Lee JK, "Digital archive system for radiologic images,'' Radio Graphics. 14:1119-1126, 1994. © 1994 Radiological Society of North America.

angiography (DSA), digital fluoroscopy (DF), and projectional radiography. These images generally are archived digitally or in analog format on different types of storage media such as magnetic disks or tapes, compact discs (CDs), optical disks, videotapes, films, digital versatile discs (DVDs), or digital linear tapes (DLTs). Retrieving these images from their archived media requires certain manually operated procedures, which tend to be tedious and inefficient. Computer-based medical image archiving was initially introduced in the early implementation of the picture archiving and communication systems (PACS) [2,3]. In these implementations, an image archive subsystem was built into a PACS, providing a reliable and efficient means to store and manage the high-capacity medical images in support of the entire PACS operations.

Implementation of an image archive system to support the operation of a PACS requires connectivity and interoperability between the archive system and the individual medical imaging equipment (e.g., CT scanners, MR imagers, CR systems) and PACS components (acquisition computers, archive system, display workstations, etc.). This always has been difficult

Copyright © 2000 by Academic Press.

All rights of reproduction in any form reserved.

because of the multiple platforms and vendor-specific communication protocols and file formats [6]. With the introduction of the Digital Imaging and Communications in Medicine (DICOM) standard, data communication among the imaging equipment and PACS components becomes feasible [5,8]. An image archive system based on DICOM [1,4] allows itself to serve as an image manager that controls the acquisition, archive, retrieval, and distribution of medical images within the entire PACS environment.

2 Medical Image Information Model

A medical imaging examination performed on a patient can produce multiple images. These images, depending on the specific imaging procedure undertaken, are generally organized into studies and series. A study is a collection of one or multiple series of images that are correlated for the purpose of diagnosing a patient. A series, on the other hand, is a set of images that are produced with a single imaging procedure (i.e., a CT scan).

Medical images generated by digital imaging equipment of various modalities are stored as information objects composed of pixels. A pixel is a data element that contains the gray level of a gray-scale image, or the RGB (red, green, and blue) value of a color image. The gray level can range from 0 to 255 (8-bit), 0 to 1023 (10-bit), or 0 to 4095 (12-bit), depending on the procedure taken by the imaging equipment. The RBG value is composed of the red, green, and blue elements, each being represented by an 8-bit value ranging from 0 to 255.

Most sectional images (CT, MRI, US, etc.) are two-dimensional. The size of such an image can be measured in terms of number of lines (rows) per image, number of pixels (columns) per line, and number of bits (bit depth) per pixel. Thus an image specified at 2495 x 2048 x 8 bits indicates the image is composed of 2495 lines, each line consisting of 2048 pixels and each pixel containing a maximum value of 255. If expressed in bytes, the aforementioned image has a size of 2495 x 2048x 1 = 5,251,072 bytes, or 4.87 Mbytes, approximately. However, for those images that consist of 10-bit or 12-bit pixel data, each pixel will require 16 bits, or 2 bytes, of computer storage, as all computers store information in units of 8 bits, or bytes. For this reason, the size of a 512 x 512 x 12-bit CT image is equivalent to 512 x 512 x 2 = 524,288 bytes, or 0.5 Mbyte. Table 1 shows the size of medical images produced by some common modalities.

3 Medical Image Archive System

The three major subsystems constituting a PACS are acquisition, archive, and display. The acquisition system comprises multiple computers to acquire medical images that are generated by the individual image modalities in the clinic. The archive system consists of a host computer equipped with mass storage for archiving the high-volume medical images in support of future retrieval. The display system comprises multiple display workstations, each composed of a control computer and image display device (e.g., high-resolution monitor) that allow a clinician to display and manipulate images [7].

The acquisition, archive, and display systems are connected to a communication network [9]. Medical images acquired by the acquisition computers from the image modalities are transmitted to the host computer (archive server) of the archive system, where they are archived to the storage devices and distributed to the appropriate display workstations.

An archive system for PACS consists of four major components: an archive server, a database management system, a storage subsystem, and a communication network. Figure 1 shows the configuration of a centralized archive system widely adopted by most PACS systems. However, a distributed archive system based on a many-to-many service model is more suitable for use in a health-care enterprise integrated PACS. The implementation of a distributed archive system is more complicated than that of a centralized archive system, in terms of system configuration and software control. The technical details comparing a centralized and a distributed archive system are outside the scope of this chapter.

Sections 3.1 to 3.4 describe the four major components (archive server, database management system, storage subsystem, and communication network) of an archive system.

TABLE 1 Size of medical images produced by

some common modalities

0 0

Post a comment