Picture Archiving and Communication System

A Picture Archiving and Communication System (PACS, such as image storage and communication system ) is in medicine a picture archiving and communication system based on digital computers and networks. The first PACS developments began in the 1970s. Significant spread in hospitals and medical practices found it, however, until the late 1990s.

PACS systems [ nb 1] acquire digital image data of all modalities in radiology and nuclear medicine. In principle, images from other imaging techniques, such as endoscopy from, cardiology, pathology and microbiology, for the PACS - ins are possible.

Individual computer systems that are permanently connected to a single diagnostic device and meet PACS tasks are referred to as mini- PACS.

Description

PACS is composed of the PACS server to which a short-term and a long-term archive is connected. The PACS server sends to viewing and Nachverarbeitungsrechner, but also communicates with its associated imaging modalities. In most cases, moreover, also a connection to the radiology information system (RIS ) is taking place. Larger PACS installations may possibly also consist of several coupled over long distances servers and archives exist.

To enable the integration of the various components to each other and the embedding of PACS in hospital information systems, DICOM and HL7 standards have been developed by international consortia. The IHE (Integrating the Healthcare Enterprise ) is an organization that brings together different standards to so-called application profiles. PACS may then correspond to one or more of these profiles.

DICOM

The most important prerequisite for the establishment of PACS systems was the DICOM standard, because by a uniform DICOM PACS server communication, and imaging devices can be used independently of the manufacturer and the connection of a device is simple and inexpensive. Modern Large Kitchen Appliances medical imaging such as CT, PET / CTs, MRs or SPECT cameras provide image data consistently in digital form according to the DICOM -3 standard. Similar to the Exchangeable Image File Format is an image or a series of images here of two parts: In addition to the actual image a wealth of information are stored in a DICOM header. These are inter alia the identity of the patient, examination date and time, clinical question, kind, type and manufacturer of the equipment used, but also the name and address of the examining institution. If present images must be recorded as film footage, these are digitized using a scanner that will get the information for the DICOM header from the RIS. Then the image is transferred to the PACS.

Especially with older devices, the standard was sometimes not adhered to or does not have all the fields filled with information, which led to communication or memory errors. Often, the ability to store the image data in DICOM standard, offered by the device manufacturer also only optional ( overpriced ). To-date ( 2011) there are imaging devices such as ( OP ) microscopes or endoscopic devices that do not provide their image data in DICOM standard. In this case, the image information can be acquired via frame grabber cards and converted to DICOM format with the help of special software products.

The functionality to convert non- DICOMBilder to DICOM or save without conversion, but increasingly also offered by PACS manufacturers and mapped in the archive store.

Servers and storage

Core of any PACS system is a PACS server. All modalities of a PACS environment to deliver their images from here and here they are also stored. In virtually every modern hospital in industrialized countries all image data are stored in the radiology PACS. In a typical 400 -bed hospital that amount of data is approximately one to three terabytes per year. The radiology archive of a university hospital may thus be several 100 terabytes in size. The exact size and quantity of the resulting images, however, varies depending on the type and number of connected modalities. Thus, a modern 64-slice CT produces a multiple of the images that outputs an older 4- slice CT. The data amount of a mammogram image is also significantly greater than that of a conventional X-ray.

In memory of the PACS archive, the image data are sometimes no longer available in the DICOM data structure. In some systems, the PACS server accepts DICOM data contrary, separates header and image data and stores both types of information - sometimes compressed - in a common database. Also additional information, such as changes or shifts of the image are stored there in addition to the information from the DICOM header. If the images are retrieved from a remote site, they are reconverted back into the DICOM format for shipping.

Data Security

Since the PACS server provides the image data for the entire institution, its failure means that no images archived and - except for the current images on the imaging device itself - and can not be viewed. Thus, the PACS must be designed not only with a high storage capacity, but also with high reliability.

Usually the image archive is divided into a short-term and a long-term memory. In the long-term memory images can be found that are older than the specified by the administrator of the system time - typically six months to two years - are. In this way, the long-term storage can be implemented more cost-effective than the short term memory. The short-term memory is designed so that it can be accessed very quickly to these frequently retrieved image data. In older PACS systems very slow tape drives or CD jukeboxes were used for long-term storage times. Today ( 2012) but are used in most cases for short-term and long-term archiving RAID archives. The long-term archive is ideally designed as RAID 61, while the short-term archive is often a RAID 51.

State of the art today (2011) are two spatially separated locations and mirrored via high -speed fiber optic connection ( > approximately 4 Gb / s ) coupled RAID systems and HA cluster. Each individual mirror consists of an independent fully functional PACS, but is designed itself also redundant. Thus, when a disk fails, an immediately available spare disk is activated, the so-called hot-spare. Network components such as Switches or fiber optic lines are duplicated. If the error is so severe that the system is no longer functional, is activated without interruption, the mirror server.

Server and RAID controllers have two to three power supplies, so that the system continues to run even when power supply failure or a circuit. Ideally, both the redundant components of a server as well as the mirror of separately fused circuits are powered and connected to an uninterruptible power supply.

In the system, accumulating errors trigger fully automated e -mails to the administrators of the system, which can thus take appropriate measures to correct them without delay. Despite these precautions, the image data are usually additionally secured to tape drives so that in the very unlikely event of total failure of the storage system or backups of the images are available.

In RAIDS in which the disks are clearly operated on the basis defined by the manufacturer to calculate the MTBF period of two to three years, it is possible to lose the data from a RAID despite the mechanisms described. On image data that are older than about two years, is rarely accessed. This means that the plate load on the long-term archive is usually very low. The drives of the RAID could therefore age or wear out, without this shows a failure of one disk. This creates the false impression that the disks in the RAID would be alright, even though they are no longer possible. If one disk fails then it is started by the RAID controller using the hot spare to rebuild the failed disk. However, the reconstruction of a failed disk is a very heavy burden on the remaining disks; thus triggered by the construction load on the remaining functioning disks lead to a failure of another drive before the rebuild is complete. In such a case, since the data in a RAID 5 would be lost, long-term archives are usually run as RAID 6. A RAID 6 can sustain even the simultaneous failure of two disks. The reconstruction of defective plates can easily take up to 10 hours. Using a RAID 6, the risk of a total loss of data in the scenario described above may be small, but not completely negligible. This is one reason why it makes sense to swap the disks in the RAID of the long-term archive on time, even if they are not defective as well as to implement the long-term archive in the form of two spatially separated RAID 6.

Workstation for image viewing and post-processing

On special workstations investigations are retrieved. While the communication between the PACS archive and the medical device follows the DICOM standard data between the workstation and the archive are sometimes transmitted in proprietary formats. The reason for this is that a network communication via DICOM query / retrieve or store is not very effective, ie there are also transmitted this much information, some of which are redundant and / or irrelevant to the image representation. Similarly, when a transfer under DICOM 3 images in a series are sequentially transferred while in a transmission via eg HTTP any image of a series can be sent first, which shortens the loading time. Therefore, the software of the workstation and the archive in general from the same manufacturer. For an HTTP -based transmission according to DICOM standard has been around a few years, the WADO standard.

Be pictures if needed, digitally enhanced in the presentation: usually the assignment of measured values ​​( X-ray absorption, signal intensity, etc.) becomes gray values ​​manipulated ( windowing ) or subsequent structure measurements performed. After reviewing the images in the light of the clinical history, the radiologist creates a written report of findings. For this purpose, the PACS workstation also a RIS client and a voice recognition software is usually installed.

On less expensive equipped workstations in the ward and clinic area pictures and report of findings can also be seen. These is usually a conventional PC; image viewing then takes place either by means of a small, special application or the web browser.

Networking with other IT systems

With the first PACS systems was planned, to link them closely with other IT systems, but initially achieved in the absence of relevant communication standards only to a limited extent. Due to the constant development of DICOM and HL7 it is today (2011 ) is usually very possible, HIS, RIS and PACS to dovetail closely.

Example: A radiological requirement that an employee logs in the HIS is passed on to the RIS, linking the survey data with the PACS studies. So it is today (2011), for example, possible that the value stored in the RIS radiological findings are available along with the PACS images in the HIS and thus throughout the hospital. The close RIS - PACS gearing allows also the call of PACS image data from HIS and RIS. The doctor selects only the patient and perhaps the testing, the PACS immediately reveals the associated image data. Also, error corrections are simplified by the close integration of the systems. If the name of the patient accidentally misspelled on registration at the HIS, so triggers a correction of patient data in the HIS from an HL -7 message that is passed to both the RIS, as well as to the PACS. A name correction " Maier Mayer " must therefore only be carried out once in the HIS, RIS and PACS are automatically synchronized.

Benefits

In contrast to the picture documentation on paper or film makers PACS systems use digital image data. This extensive opportunities arise to increase the functionality and efficiency of work processes.

The digital storage, the quality of the recordings remains unchanged for many years. In projection radiographic methods, the digital recording allows a higher dynamic range. Recordings are thus more informative, retakes after miscarriage exposures less common than in the film radiography.

For cross-sectional imaging is expanded opportunities for observation and diagnosis result. Thus, an average series are displayed as an animation or anytime reworked in a MPR or in a 3D model converted or with special evaluation. PACS also simplifies the documentation of moving images at ultrasound.

A major advantage is the simultaneous availability of images at multiple locations (even within a hospital ) over a computer network, which can account for the logistics of the conventional image transport completely. Since the images can be reproduced over long distances, the assessment can be made more flexible in time and space (see also teleradiology ). Recordings can be copied without loss. Cumbersome Filmarchivierung omitted. The risk of loss of unique original recordings is minimized.

Savings in visual media, transport costs and filing space are other important advantages.

Disadvantages

In the first years of the introduction of PACS PACS environments diseased often to poor DICOM implementations, so that seemingly DICOM compatible devices often could not be tethered or data were limited memory or read. Also, the computer and network architectures of the first PACS servers and workstations the size of the image data set were not up, which led to excessively long transfer and charging times. A high maintenance, high system prices and sometimes low reliability resulted in PACS and the PACS concept often heavily were criticized and the benefits of PACS has been questioned.

Despite great effort and hardware- redundant servers, it is possible that the PACS can not be accessed. Causes may be: Physical interruption of network connections, failure of switches, misconfiguration, for example, recently admitted into the network devices or the crash of server services. Such an interruption usually affects several, in the worst case, all users of the PACS. As with other central, server- based IT systems, the resulting loss of productivity is therefore usually high.

History

Although medical image data were digitally stored already in the 1970s, were attempts to establish digital archives, proprietary, in the absence of standardized data formats island solutions. First begun in 1983, development of the DICOM standard also allowed for a multi-vendor PACS development. As the first devices that could output their data in DICOM format, with the adoption of the DICOM 3.0 is still valid today only came in 1993 on the market, PACS systems were hardly to be found towards the end of the 20th century in hospitals. Still in 2005 had only an estimated 22 % of all North American hospitals a PACS system.

With adoption of the DICOM standard in 1993, the data transfer rate of Ethernet was 10 Mbit / sec. Even high-priced workstations had a few dozen megabytes of RAM and only a few hundred megabytes comprehensive hard disk drives. Today (2011 ), the continuous transfer rate of Ethernet at 1000 Mb / sec. A low-cost PC "off the shelf " has several gigabytes of RAM, hundreds of gigabyte hard drives and higher by several orders of magnitude processing power than computers from the 1990s. The amount of image data in radiology is even increased in the course of about twenty years PACS development, but not nearly to the extent that the increased performance of computers and network infrastructures. A chest X-ray is now the same as 20 years ago. In combination with the ongoing price decline of IT systems and further developed as well as better -observed communication standards, this led in recent years to a very high prevalence of PACS. The benefits of PACS outweighs now clear the associated costs.

Classification as a medical device

PACS software (eg for archiving and reporting of image data ) is brought into the European Union in general as a medical device in Class IIa in traffic. If the PACS software affect the operation of a medical device has associated with it (eg functionalities for radiotherapy planning ), it can be Class IIb. The conformity assessment procedure is regulated by a so-called Notified Body. Medical products such as PACS consist possibly of several components that can be evaluated individually. These components can, therefore, also have a function of their respective purpose, different classifications.