| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| ||||||||||||||||||||||||||||||||||||||||||||||||||||
Commentary |
ImPACT, Bence Jones Offices, Perimeter Road, St George's Hospital, London SW17 0QT, UK
The development of CT scanner technology in the past 10 years has brought a range of advances, including slip ring scanning, fast (sub-second) rotation times and greater computing power. These advances have been combined to allow rapid reconstruction of CT scanner data that can be updated continuously to provide cross-sectional images in near real-time. CT scanning is conventionally prescriptive, with scan sequences planned in advance. Real-time scanning expands the role of CT to an interactive imaging technique.
Applications of real-time CT lie in two main areas: timing and monitoring of helical scan sequences, and interventional procedures such as tissue biopsy and the drainage of fluid-filled lesions. In this paper, the use of real-time CT in interventional procedures is referred to as CT fluoroscopy. With the advantages of this new technology comes concern over doses to both patient and equipment operator, as well as the need for careful control over its use.
This commentary examines the technology, applications and dosimetry of real-time CT scanning and draws out some of the important points in each of these areas.
Technology
In 1993, Toshiba Medical Systems Co. Ltd (Tokyo, Japan) began the development of real-time capabilities for their Xpress/SX CT scanner [1]. This scanner had a gantry rotation time of 1 s, and the Aspire CI real-time package was capable of reconstructing and displaying three images per second. Each image displayed on-screen represented the object in the scan field of view in the preceding second. This did not match the almost instantaneous display provided by conventional planar fluoroscopy units, but was rapid enough to provide feedback on what was currently in the scan plane and allow the scanner operator to react to that information.
Subsequently, all the medical CT scanner manufacturers have developed real-time scanning systems. These have generally been for their top-of-the-range single slice scanners, but real-time CT capabilities are also being developed for multislice scanners, with the potential to display more than one adjacent real-time image simultaneously. Despite minor differences, the manufacturers' systems have a lot in common. Scanning is generally performed at 120 kV with tube current in the range 30–90 mA and a scan time of 1 s and below (down to 0.5 s for the fastest system). Images are displayed at a rate of between 3 s-1 and 12 s-1. In general, a scanner with a fast tube rotation time will provide better feedback to the operator than a scanner with a more rapid image display rate and a slower tube rotation time.
For interventional procedures, images are displayed on a monitor in the scanner room. Exposure is controlled with a foot switch, and couch movement is achieved either with a joystick mounted at the side of the patient table or using a freely floating couch top that can be moved by hand. "Last image hold" retains the final image in a sequence on the monitor after scanning stops.
Reconstruction times for real-time CT scanning seem very fast compared with the 1–2 s per standard image for a modern single slice CT scanner. Rates of three images per second and above are achieved using a number of simplifications in the reconstruction process. The real-time image is generally reconstructed on a 256 x 256 pixel matrix, rather than the more usual 512 x 512 pixel matrix for conventional CT. Other image calculations, such as beam hardening corrections, may also be omitted. Image display rates are also helped by the fact that, in order to display the nth image in a sequence, only the projection data not present in the nth-1 image needs to be back projected. As display rate is increased, the number of calculations required to produce each image decreases.
All CT fluoroscopy systems on the market have an alarm timer, similar to those in conventional planar fluoroscopy. There is no standard time limit for these alarms, but a 100 s limit is pre-set on a number of models.
Real-time CT and CT fluoroscopy are often sold as separate packages, so purchase of a system with real-time capabilities does not necessarily imply CT fluoroscopy facilities will also be included. Upgrade options are available for most modern single slice scanners, usually consisting of exposure foot pedal, table top control mechanism and in-room monitor, and sometimes including hardware to enable rapid image reconstruction.
Applications
The ability to visualize cross-sectional CT images in real-time has been applied to two main tasks: timing of beginning and end points for helical scan sequences, and interventional procedures (CT fluoroscopy).
Scan timing
The time between injection of CT contrast medium and its appearance in the organ of interest varies from patient to patient. Producing optimal contrast studies requires images to be acquired at the correct vascular phase. Real-time CT can be used to aid this process by continuously monitoring a region of interest, such as the aorta, and starting the conventional helical run once the mean CT number in this region reaches a pre-set threshold [2]. This process is commonly known as bolus tracking and has been shown to be effective in increasing contrast enhancement [3, 4].
Real-time CT scanning can also be employed to offer a continuous transaxial view during helical scanning. This allows the operator to stop acquisition before the prescribed end point, for example if the patient moves significantly or if the entire organ of interest has been imaged before the end of the scan run. For a helical run planned from a scan projection radiograph the latter situation should not occur, so the usefulness of real-time helical scanning is generally limited to observing images rather than reacting to them.
CT fluoroscopy
The use of the term "CT fluoroscopy" varies, but in this article it describes application of real-time CT scanning in interventional procedures. Images from the scan room monitor enable the operator to guide a needle to a specific site within the body. Silverman et al [5] described two main approaches to this. The first uses continuous imaging and a needle holding device, similar to sponge forceps, to keep the operator's hands out of the primary radiation beam. The second employs intermittent imaging and manual needle manipulation. Of these, the first can result in higher doses to patient and operator but provides a less tactile method for needle placement. The second is closer to existing methods for non-real-time CT biopsy but has the advantage of shorter periods between scanning and adjustment of needle position, which improves the interactivity of the procedure.
The main reported applications of CT fluoroscopy [5, 6] are tissue biopsy and drainage of fluid-filled lesions, as well as a variety of other procedures such as ethanol ablation of tumours, placement of catheters and guidance of sacroiliac injections.
Dosimetry
The dosimetry concepts in conventional and real-time CT are broadly similar but bring different issues to the fore. For most real-time applications, scanning takes place in one scan position, meaning that local patient skin doses have the potential to reach levels where deterministic radiation effects are seen. In CT fluoroscopy the equipment operator can receive considerable doses. This is a departure from the usual situation in CT scanning, during which the equipment operator is located in a separate control suite and receives negligible radiation dose.
Published values of screening times for CT fluoroscopy procedures [5–8] range from 3 s to 660 s, with mean screening times in these studies of 79–165 s. This wide range reflects differing techniques, levels of operator experience and workloads studied, as well as the degree of complexity of individual procedures.
Dose to patient
A recent ImPACT "blue cover" report [9] examined the dose from abdominal CT fluoroscopy procedures using typical exposure parameters (120 kV, 50 mA, 1 s rotation time, 10 mm slice thickness) for a range of scanners. Skin dose rate and effective dose rate were estimated to be approximately 4–5 mGy s-1 and 60 µSv s-1, respectively. For the previously discussed studies, this would result in mean skin doses of 0.4–0.8 Gy. The study with a maximum scan time of 660 s would result in a skin dose of approximately 3 Gy. This is greater than the threshold level for transient erythema (2 Gy) and in the region of that for temporary epilation (3 Gy). Mean effective patient doses would be between 4.7 mSv and 9.9 mSv for these studies. This compares with an effective dose of 11.7 mSv for a standard abdomen scan using the "European Guidelines on Quality Criteria for Computed Tomography" [10].
The above values are examples of the magnitude of patient doses. When studying individual patient doses, it is important to use actual exposure parameters and scanning time in dose estimations as there is potential for wide variation from patient to patient.
Dose to operator
In addition to operator position relative to the scan plane, the variables relevant to staff dose are the same as those influencing patient dose (i.e. kV, mA, slice width, total scan time). Minimizing operator dose can be achieved by keeping as far from the scan plane as possible.
Estimates of dose rates to the operator from CT fluoroscopy drop very rapidly with distance from the scan plane. ImPACT measurements on a Toshiba Asteion (Toshiba Medical Systems Co. Ltd, Tokyo, Japan), with a 32 cm CT dose index phantom to provide scatter were 4 mGy s-1 in the scan plane, 9 µGy s-1 at the body trunk and 2 µGy s-1 to the eyes (operator positioned 40 cm from the scan plane). These measurements are obviously sensitive to operator position, but correspond well with those of Ozaki [11].
It is difficult to estimate dose per procedure to staff in the scan room, as operator technique is a key factor. However, careful monitoring of doses to hands, trunk and eyes is needed to ensure occupational dose limits are observed. Use of lead aprons by the operator is strongly recommended; Ozaki noted that they reduced dose rate to the trunk by a factor of 14. Consideration should also be given to the use of thyroid shields and lead glasses.
Conclusions
Real-time CT and CT fluoroscopy are useful techniques offering a number of potential benefits, including more consistent timing of contrast studies and improved diagnostic accuracy of biopsies. As with other interventional radiology techniques, doses to both patient and equipment operator from CT fluoroscopy have the potential to reach high levels and should be monitored carefully. In particular, an intermittent imaging technique is preferred to continuous imaging using a needle holder owing to reduced radiation doses and improved needle manipulation.
The availability of CT fluoroscopy is still limited, but as its use becomes more widespread it is likely to become a routine aid to a growing range of procedures. Biopsy guidance is the most obvious application of this technology, but any technique requiring real-time three-dimensional guidance of a probe through the body could potentially benefit.
Footnotes
ImPACT is the UK's CT scanner evaluation group, funded by the Medical Devices Agency. 
Received for publication October 31, 2000. Revision received May 31, 2001. Accepted for publication June 11, 2001.
References
This article has been cited by other articles:
 |
|
 |
|
  R. M. S. Joemai, D. Zweers, W. R. Obermann, and J. Geleijns Assessment of Patient and Occupational Dose in Established and New Applications of MDCT Fluoroscopy Am. J. Roentgenol., April 1, 2009; 192(4): 881 - 886. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
D. D. Brennan, L. Appelbaum, V. Raptopolous, J. B. Kruskal, and S. N. Goldberg CT Artifact Introduced by Radiofrequency Ablation. Am. J. Roentgenol., May 1, 2006; 186(5 Suppl): S284 - S286. [Full Text] [PDF]
|
|
|
|
 |
|
 P. Aviles Lucas, D. R. Dance, I. A. Castellano, and E. Vano Estimation of the peak entrance surface air kerma for patients undergoing computed tomography-guided procedures Radiat Prot Dosimetry, May 17, 2005; 114(1-3): 317 - 320. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 P Dawson Patient dose in multislice CT: why is it increasing and does it matter? Br. J. Radiol., December 1, 2004; 77(suppl_1): S10 - S13. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| BJR | DMFR | IMAGING | ALL BIR JOURNALS |
