American College of Cardiology

The primary technical focus in the cardiac catheterization laboratory is the generation, recording, and display of high-quality x-ray images during diagnostic and interventional catheterization procedures. With greater numbers of increasingly complex interventional procedures being performed, acquisition of fluoroscopic and cineangiographic images of the highest quality remains crucial for the optimal performance of the catheterization laboratory. The ongoing trend toward more complex interventional procedures results in greater exposure to radiation for the patient and laboratory staff ⁽⁹³⁾. The longer procedure times associated with these procedures also place greater demands on the x-ray generator and tube than may have been the case previously. During the last half-decade, the prominent role of 35-mm cine film as the recording and archiving medium has been challenged, and cine-less operation has become accepted as routine practice in many laboratories ⁽⁹⁴⁾. The movement toward digital technologies in the catheterization laboratory and throughout the hospital and community continues to change the interaction between the laboratory and the outside world. The rapid evolution in information technology in turn changes users' expectations regarding the accessibility of medical image data ⁽⁹⁵⁾. These significant technical changes require that the basic requirements in the catheterization laboratory be revisited with some frequency to ensure that the technical needs of the laboratory are being met appropriately.

A. Radiographic Equipment

The conflict between acquisition of high-quality angiographic images and limiting x-ray exposure of patients and staff has always been difficult to resolve satisfactorily. Cardiac angiography, with its simultaneous requirements for high acquisition rates and the need to visualize very small objects, places some of the most severe demands on x-ray generating equipment. Although there has always been optimism that improvements in technology would help resolve the conflict, the increasing clinical requirements instead have led to yet greater demands on the equipment. The wide acceptance of digital angiographic systems, which have advantages in a variety of procedures, has led to new challenges and concerns about issues of data rates, amounts, and storage ^(96,97). High-quality video display has become standard in the laboratory, and the use of pulsed-progressive fluoroscopy is assumed with any currently available equipment ⁽⁹⁸⁾.

B. Generators

The rising proportion of interventional procedures performed in the catheterization laboratory increases the demands placed on the x-ray angiographic equipment, including the x-ray generator. These requirements typically include a high-frequency generator with outputs of 80 to 100 kW at x-ray pulse rates of 30 pulses per second (60 pulses per second for pediatric applications). The ability to perform pulsed fluoroscopy and angiography with short exposure times—sufficient to avoid motion blur of objects moving at high speeds but still able to maintain a high degree of contrast—has become a minimum requirement in angiography. Modern equipment designed for the catheterization laboratory should also have automatic exposure control (AEC), which provides the optimal combination of x-ray tube voltage, current, and exposure time most suited for visualization of rapidly moving coronary arteries with adequate contrast. Many laboratories choose to use the high-level control (HLC) fluoroscopic technique, which can produce exposure rates beyond the standard regulatory limit of 10 roentgens per minute (R/min) but less than that used during cineangiography. There are broader implications for this higher exposure mode related to the resulting exposure to the patient and staff, but if the capability is deemed a requirement for a particular laboratory, a generator with such capability will be required regardless.

C. X-Ray Tubes

Along with the high-output generator, the x-ray tube used in cardiac catheterization laboratories—especially for long, complex interventional procedures—must meet the most demanding technical requirements ⁽⁹⁹⁾. To visualize the smallest coronary arteries and complex variations in lumen geometry, focal spots of 0.6 to 0.8 mm are necessary. To acquire multiple angiographic sequences lasting up to 10 to 30 s at rates of 30 frames per second (and higher), the tube must be able to absorb large amounts of energy and dissipate the resulting heat quickly to avoid delays between acquisitions—and serious damage to the x-ray tube. Similarly, the extensive fluoroscopic times required for interventional procedures will be limited by the heat dissipation characteristics of the x-ray tube. Heat storage capacities >1 million heat units (HU) and even approaching 3 million HUs have been found invaluable in the catheterization laboratory. A tube with greater heat storage capacity provides several significant advantages in the modern clinical environment: (1) it reduces delays during clinical procedures if lengthy fluoroscopic and angiographic exposures result in the heating of the tube to its maximum capabilities; (2) it allows for penetration of larger patients (or at steeper angulations) without resorting to higher x-ray tube energies and the resulting reduction in vessel (and device) contrast and increase in image noise; (3) it provides the option for use of filtering materials that can either reduce patient skin exposure, improve image noise characteristics, or both.

D. Image Intensifiers

The performance characteristics of image intensifiers used in the catheterization laboratory have improved continuously over the years. Higher conversion factors—greater efficiency in light output as a function of x-ray input exposure—along with improved contrast ratios and a reduction in geometric distortion have led to improved image quality. As a result, image intensifier performance—in dedicated cardiac catheterization systems—has been optimized for the task of imaging the heart and coronary arteries at the x-ray exposure levels used in that application. In general terms, this translates to a high-contrast spatial resolution >3.0 line pairs/mm at the entrance plane of the image intensifier and acceptable signal-to-noise quality at cineangiographic entrance exposures of 20 to 25 microR/frame in the typical magnification mode used for coronary angiography ⁽¹⁰⁰⁾. On the other hand, systems optimized for other imaging tasks, such as peripheral and vascular angiography, cannot at the same time deliver the required performance in the heart, and users should be aware that there will be degradation in performance for the task of imaging the heart and coronary arteries. The different x-ray tube design, e.g., target size and angle, required for covering a field of view corresponding to a 40-cm image intensifier will not deliver the same results at the same x-ray exposure for the 12- to 20-cm field of view customary in cardiac imaging. Similarly, the light-gathering characteristics of the larger image intensifier may compromise performance for cardiac imaging at the smaller fields of view when only a fraction of the intensifier input surface is used. In turn, there are implications for the electrostatic optics within the intensifier as well as for the light-gathering optics at the output surface of the intensifier; in general, these effects will degrade the image quality relative to an intensifier specifically designed for the cardiac application. In summary, the effects that can occur in such a system are degradation in spatial resolution, increased image noise in fluoroscopy and angiography, increased x-ray exposure rates, and delays due to exceeding heat capacity for x-ray tubes.

E. Developments in X-Ray Detectors

As mentioned above, the image intensifier is a vital component for x-ray angiographic imaging in the cardiac catheterization laboratory, because no better alternatives exist (to date) that can convert the x-ray intensity information exiting the patient into usable, visible light information. This visible light is converted to an electrical video signal and in turn to a stream of numbers for subsequent processing and storage. Recent developments in digital x-ray detector technology demonstrate significant potential for application in cardiac fluoroscopic and angiographic applications in the future. The first commercial systems have only recently been introduced, and it is expected that they will become more common; readers should thus be aware of how these systems differ from those to which they have been accustomed. The most significant difference between the image intensifier-based imaging chain and those based on digital detector technologies is that the entire image intensifier tube will disappear, along with the video camera used to convert the output light signal to an electronic voltage. The promise is that this will deliver significant improvements in spatial resolution because it will not be limited by "blurring" processes inherent to the image intensifier systems, greater dynamic range and contrast resolution, and because the new detectors are much simpler mechanically, with less likelihood of failure and gradual degradation in performance ⁽¹⁰¹⁾.

The candidate detectors that will most likely be appearing in cardiac catheterization laboratories are digital “flat-panel” detector systems that use a compact box in place of the image intensifier tube to convert the incident x-ray signal directly, essentially to an array of discrete electrical signals that are read individually and processed, displayed, and stored for further processing and review. These types of detectors are characterized as direct digital because the intensity values are generated as an array of digital values (ones and zeros) at the detector without the need for additional conversion processes that can add noise or degrade performance ⁽¹⁰²⁾. These detectors have been made possible by improvements and cost reductions in the manufacture of flat-panel displays used in the computer industry (e.g., laptop computer monitors). Essentially, these active matrix arrays are combined with an x-ray-sensitive layer that converts the x-ray signal to light incident on an array of light-sensitive cells, anywhere from 1000X1000 to 2000X2000 cell arrays. Clinical testing of the first of these devices has begun, and these detectors are now becoming available as product options.

F. Video Components

1. Video Cameras

Along with image intensifiers, high-quality video cameras have been an assumed component of modern cardiac angiographic systems for generation of high-quality images during fluoroscopy and to provide the analog signal source for the conversion process used in all current-generation digital angiographic systems. The traditional "pickup tube" camera, based on a scanning electron beam to read off the spatially varying electrical signal produced by the light output of the image intensifier, is well understood and has been described in detail in the previous guidelines ⁽⁵⁾. Modern systems use cameras that operate in "standard" resolution mode (525 lines per video frame) as well as "high" resolution mode (1023 or 1049 lines) for both fluoroscopic and angiographic applications. An important characteristic that should be carefully assessed in these systems is the introduction of additional electronic noise to the image in the higher-resolution modes requiring higher bandwidth electronics. Another aspect related to the discussion of "dual-use" systems above relating to image intensifier performance is the fact that video systems designed for slow frame rate angiographic applications (typically at higher x-ray entrance exposures) may demonstrate degraded temporal performance (i.e., blurring) when operated at the faster acquisition rates required for cardiac imaging.

Although the pickup tube video camera has long been the workhorse of cardiac angiographic systems, a relatively recent development has been the increasing availability of video cameras based on solid-state image sensors (e.g., the charge-coupled device or CCD cameras) ⁽¹⁰³⁾. CCD sensors consist of an array of discrete elements (typically 1024 X 1024) that store the light information from the image intensifier output until they are read by the camera electronics. Among the advantages of the CCD camera are simpler design, resulting in smaller size; improved dynamic range; improved spatial resolution; absence of temporal lag; prolonged life; lower cost and simplified maintenance requirements; and lack of sensitivity to magnetic fields. (Note that the flat-panel detectors described above also share a number of these advantages.) The advantages listed above have been extensively demonstrated and, in general, CCD cameras operating at the same pixel resolution do offer better image quality than video tube cameras. It should be understood, however, that the use of CCDs in the detector chain does not itself make this a direct digital detector. In most applications, the output of the CCD camera is a time-varying analog voltage signal that must be digitized before use as with pickup tube cameras. One issue that should be considered is the increased amount of data that can be generated by the larger matrix sensors.

G. Digital Angiography Issues

Among the most significant changes in practice in the cardiac catheterization laboratory in recent years has been the consistent move away from 35-mm cine film as the standard recording medium. In many laboratories in the United States as well as abroad, cine film is no longer used in any function of the catheterization laboratory. This has advantages that have been well documented ^(103,104), but there are also a number of changes in the technical requirements for a laboratory and considerations that must be made in equipment purchase and use.

H. Effects on X-Ray Requirements

For many years, the promise of digital angiographic recording has been accompanied by the promise of reductions in x-ray exposure ⁽¹⁰⁵⁾. This has in turn led to confusion when this reduction has not materialized as digital angiographic systems are implemented. The image quality in cardiac catheterization specifically at the exposure per frame that is customarily found in this application is primarily a function of the contrast signal and the image noise that are found acceptable for the application. With state-of-the-art equipment, the primary source of image noise resides in the image intensifier, and this may be objectionable. The light output of image intensifiers designed for cardiac catheterization is more than enough at standard entrance exposures for both cine film and digital recording. Elimination of the cine camera does not by itself mean that more light is available for the video chain; there are already adequate amounts of light to obtain good video signals. However, the noise is directly related to the statistical properties of the relatively low x-ray beam flux incident to the image intensifier. It is true that a digital-only angiographic system allows more flexibility with the light-limiting apertures in the video chain because one is not limited by the requirements of delivering adequate light to the cine camera. But the use of such apertures as a dose-adjustment method is not routine. More practically, the elimination of cine film does allow use of reduced frame rate acquisition with the proportional reduction in x-ray exposure ⁽¹⁹⁾. In the end, however, x-ray exposure per frame is affected more by other factors, such as those described above.

I. Digital Acquisition Requirements

As discussed above, at the time of this writing, essentially every digital angiographic system requires a conversion from an analog signal produced by a video camera to a string of numbers stored and processed for display and analysis. The spatial, contrast, and temporal resolution requirements for cardiac catheterization are well understood and have been met for the most part with digital systems operating at matrix sizes of 512 X 512 pixels, bit depths of 8 bits (corresponding to 256 intensity values), and acquisition rates of 30 frames per second. With the ongoing reduction in the cost of digital hardware, it has been possible to deliver improved performance as well. Newer digital systems can record to larger image matrices (1024 X 1024 is common) and greater bit depths, with 10- and 12-bit images becoming available. With pickup tube cameras, higher spatial resolution is a function of the camera scanning rate along with the image intensifier magnification mode (and geometric magnification). With CCD cameras, the size of the element on the camera is necessarily fixed, but improvements in resolution are still possible through the use of higher magnification modes, which map a fixed-size element to a smaller object in the patient. Similar factors apply as well to digital detector technologies, but in that case, the resolution cannot be improved through selection of a higher magnification mode. In a field of view of 15 cm—typical for coronary angiography—a 512 matrix results in a limiting resolution of 0.20 to 0.25 mm (in the plane of the imaged object), corresponding to a resolution of 2.0 to 2.5 lp/millimeter. While this has been found in general to be adequate for coronary imaging, it is less than the theoretical resolution of cine film. Accordingly, many vendors are offering higher resolution systems. In considering these systems, however, it is important to ensure that the contrast resolution is adequate as well. For instance, if there is insufficient contrast from small objects due to other factors in the imaging chain—x-ray energy, image intensifier or video issues, or radiation scatter— the improved spatial resolution will be of limited advantage. The issue of much greater data storage requirements must be considered as well.

J. Digital Storage and Display

In most digital angiographic systems in use today, there is limited storage capacity on the system itself, usually only enough for one to several days’ worth of procedures. It is therefore necessary to have a mechanism for medium- and long-term storage. The development of the DICOM standard for cardiac angiography accelerated the use of digital storage media by ensuring that there was a well-understood method for exchanging digitally recorded procedures between laboratories and systems ^(106,107). Many laboratories purchased the capability for storing exams on CD-ROMs and have pursued that as a long-term storage medium. Other laboratories have implemented automated storage libraries, which provide access to many months or years of procedures without the need for manual intervention for retrieval and display. The specific archival systems used by a laboratory should be selected on the basis of clinical and financial considerations ^(96,97).

Whatever approach is taken, every laboratory should ensure that the version of data retrieved from the archive at some later date is identical to the version used for postprocedure diagnosis and decision making. In most laboratories the digital image data are typically not reintroduced into the digital angiographic system for later review, but rather are reviewed on a separate digital image review workstation. This workstation can be supplied by either the original x-ray equipment vendor or increasingly from a variety of third-party manufacturers. The quality and cost of such workstations can vary greatly. A laboratory should ensure that review performance from media or over a network is adequate for subsequent clinical assessment. Among the factors that must be considered are display rates equivalent to the original acquisition rate, full image display resolution, image processing and enhancement, and sufficient exam storage capacity to avoid the need for delay in retrieval of older exams. It should be noted that both the technology and cost of imaging workstations are changing rapidly, and a laboratory should anticipate future needs at the time of equipment purchase.

K. Image Formats and Standards: The DICOM Standard

The acceptance of the DICOM standard for cardiac angiography provided assurance that a version of an angiographic exam exchanged with another laboratory or reviewed at a later date could be identical to that reviewed in the laboratory during and immediately after a catheterization procedure. It is unfortunately not always the case that every vendor claiming to subscribe to the standard delivers equipment that does in fact meet the functionality stated above. The DICOM standard does include a conformance mechanism by which a vendor is required to list the capabilities and supported features of the product, but the standard remains relatively new, and unfortunately, a laboratory cannot rely solely on such claims. Any archiving, exchange, or review system should be carefully assessed to determine whether the exam data stored for local archiving or written to media for exchange do in fact provide the same quality of diagnostic information achieved in the original.

One factor that can ensure such equivalence is the use of a direct digital interface between the angiographic acquisition system and the archiving or review system. Due to the relatively recent introduction of the standard and the fact that there is a large installed base of equipment of varying age, this digital interface is achieved through a range of sometimes complicated approaches or, for some of the oldest equipment, cannot be achieved at all.

Newer digital angiography equipment is available with a DICOM network interface, meaning that the exam information leaving the system is already formatted according to the DICOM standard. This has 2 advantages: (1) data equivalence is assured and (2) any receiving system that supports this interface can be used for storage and review (i.e., a laboratory has more choices). An alternative approach typically used with older equipment is a digital interface, which requires an additional step to format exam data to the DICOM standard. This approach can work as well, but it requires that the developer of the interface, usually the manufacturer of the x-ray system, make the interface specification available to other vendors. Otherwise, only equipment from the original vendor will support the interface. Again, it should be noted that either type of interface will usually work, but it should be understood from the onset which type is being provided.

Another alternative that has also been implemented, especially with older equipment, is an analog capture interface rather than a digital interface. In this approach, the analog video signal is captured and digitized somewhere in the acquisition or display chain, and this second digital version is then formatted according to the DICOM standard and used for archiving, display, and exchange. Laboratories should be cautioned that the image data resulting from this approach is not strictly equivalent to the information available at the time of the procedure. With appropriate precautions, the quality can still be quite high, but, in many cases there is significant degradation in image quality. In the case of older equipment for which no alternative exists, this approach does provide a digital archive and exchange approach of nearly equivalent image quality, but only if the parallel image capture is performed to a specification close to that incorporated within the original x-ray vendor's analog-to-digital conversion. In some cases, irreversible data compression is also used during the capture process to save costs and improve performance, but this leads to artifacts in the stored archival copy, which degrade image quality and can affect visualization of the anatomy.

L. Digital Image Resolution

As discussed earlier, digital cardiac angiographic systems have most often incorporated digital resolutions of 512 lines by 512 columns by 8 bits per sample—usually together with acquisition rates of 30 frames per second. The initial basic version of the DICOM standard for digital cardiac angiography supported only this format to provide at least 1 format that many vendors could support. In recent years the number of cardiac angiographic systems available at higher resolutions has increased, and the issue of digital equivalence between the acquired exam information and the stored and exchanged versions has emerged. The issues include the clinical requirement for the higher resolution and whether the permanently stored copy should also be stored with the higher resolution. In general, a matrix size of 512X512 has been deemed acceptable for clinical applications despite the fact that this corresponds to a minimum spatial resolution on the order of 0.2 to 0.3 mm.

For some applications, such as quantitative coronary angiography of complex stenoses, higher resolution (e.g., 0.1 to 0.15 mm) is seen as optimal ⁽¹⁰⁰⁾. In that case, the 1024 X 1024 matrix size will deliver improved resolution but at the cost of increased data acquisition, storage, and transmission requirements. Under these circumstances, the version of the exam stored locally as well as used for exchange should accommodate the higher resolution images on which the clinical diagnosis was made. The DICOM standard does accommodate these higher image matrices, but a laboratory should ensure that the equipment purchased for this application does indeed store all the acquired information and can write exchange media in the larger format (assuming that the receiving laboratory in turn has the ability to display the larger format images). Similar concerns apply to network transfer, which will require greater bandwidth capacity for the larger amounts of data or, alternatively, greater delays in transmission.

M. Data Compression

Despite the rapid improvements in digital storage and transmission technology, the data requirements of digital cardiac angiography remain among the most demanding in medical imaging. As a result, some equipment vendors and suppliers of imaging applications have sought to reduce the amount of data through the use of mathematical techniques or compression methods ⁽¹⁰⁸⁾. Some compression methods are completely reversible or lossless, and in the end the recipient or user has precisely the same equivalent data that were available initially. In contrast, irreversible or lossy compression methods reduce the amount of data, but the resulting image information is not strictly identical to the original angiographic information. The variation that can be detected visually and the effect it may have on clinical assessments made from the images depends on the type of compression method and degree of reduction. Lossy compression methods can result in much greater amounts of data reduction, but the resulting images may contain detectable artifacts that were not in the original image. The basic problem with the use of irreversible compression methods is that a user at a later time is not provided with the same information used initially; this may be acceptable in some applications, but none in which the original information is no longer available if needed. The ACC and the European Society of Cardiology have sponsored a multicenter clinical study to assess the effects of one of the more common compression methods (motion JPEG) on critical diagnostic tasks. The results of that study indicate that as the amount of compression is increased, the ability to detect clinical features is impaired. It is strongly suggested that laboratories avoid the use of lossy compression methods for the permanent storage of digital angiographic data ⁽¹⁰⁹⁾.

N. Telemedicine Applications

Routine storage and availability of angiographic records in a digital format makes possible a new class of clinical applications, which fall under the general category of telemedicine—referring to the electronic transmission of clinical image data over large distances to support clinical decision making at remote sites ⁽¹¹⁰⁾. In addition to the digital format of the procedure data in the acquiring laboratory, this requires an accepted standard format for the image data that can be displayed at the receiving site as well as a reliable digital network link between the sending and receiving centers. Such Wide Area Network (WAN) applications extend, in simplest terms, the network from within a laboratory or hospital to much greater distances. The network in effect provides a user hundreds of miles away with the ability to display and review the procedure image data as if he or she were in the procedure laboratory.

In one example of this type of application, simple "store and forward" transmission of a procedure from a referring hospital to another medical center may be performed, in effect replacing the mail or courier service with electronic transfer. One advantage of this approach is that the original copy of the exam record remains in the acquiring laboratory. At the other end of the spectrum, "expert" physicians at one center can participate in and provide advisory support in real time for a procedure being performed at another center.

Although the cost of digital network transmission is being reduced and networks are being extended to more locations, the bandwidth needed to transmit cardiac catheterization examination results rapidly and completely remains costly and/or relatively rare at this time. As with many networking applications, users must choose between speed and expense. As a result, some attempts to provide telemedicine services for catheterization procedures in a less costly manner have incorporated data compression methods to reduce the size of the data files being transmitted and, in turn, reduce the time required to transmit the files. As with the issues raised in the discussion above, this means that the data being reviewed at the distant location may differ to a clinically relevant degree from the data used in the initial review and diagnosis. There may be applications for which this is acceptable, but laboratories and hospitals should not use such systems for an application that involves decision making solely on the basis of compressed images, which are degraded in diagnostic quality. Specifically, systems developed for routine video conferencing over standard telephone lines do not as a rule have adequate image quality for transmission of clinical image data acquired under the conditions required for cardiac angiography. Similar concerns apply to transmission over the Internet using standard networks. There are technical means available to deliver diagnostic image quality of digital angiograms in a timely fashion; laboratories considering a telemedicine application should examine the technical specifications of such systems carefully. In the near future, clarification of privacy concerns for individual patients should allow for discussions of patient cases without jeopardizing patients' confidentiality.

O. Quantitative Measurement Methods

With increasing storage and easy access to catheterization results in a digital format, the use of computerized methods to measure coronary artery stenosis severity, left ventricular function, regional wall motion, and other techniques becomes much more feasible ^(111,112). These methods provide objective, reproducible measures for assessment of disease and physiological function and are available for use in all laboratories rather than being limited to multicenter clinical trials. Laboratories should ensure that systems being used have been fully validated in well documented in vitro and in vivo studies and that technicians are familiar with the procedures necessary to produce measurements that are reliable, accurate, and reproducible. With the elimination of cine film and its replacement with digital images that are readily accessible, one of the greatest sources of variability is no longer an issue with the elimination of the need to convert cine film to a digital file. Consistent procedures must be followed, and it is important that the analysis be performed on the original image data acquired at the time of the procedure. As mentioned above, systems are available that store and exchange versions of the images that differ significantly from the original acquired data; in general, results on such secondary capture images or, possibly, compressed images will not be as reliable.

P. Further Developments in the DICOM Standard

Just as the DICOM standard for digital images made it possible for catheterization laboratories to routinely exchange high-quality digital images, other developments are under way to enable similar standardized formats for other types of information gathered during catheterization procedures ⁽¹¹³⁾. An outgrowth of the DICOM effort extends the standard format to include physiological data: hemodynamic and electrocardiographic waveforms and measurement results. The primary focus of the effort was to facilitate a "unit patient record" for the catheterization procedure that includes all information necessary as a local record and for interchange between laboratories and hospitals. Although the standard is in a preliminary stage at this time, newer versions of imaging and waveform recording equipment will support the standard, and laboratories that wish to incorporate such capabilities should obtain assurance that all required equipment supports the standard. In a related development, standard formats can ensure that the results of the procedure data can be exported electronically in a format compatible with national data registries such as the ACC-NCDR™ ⁽¹¹⁴⁾. Other developments include the integration of other imaging modalities such as intravascular ultrasound, as well as standard formats for clinical reports that can be transmitted electronically in such a manner that they can be displayed in other laboratories regardless of which vendor's equipment was used to generate the original procedure report.