2nd CT users group meeting: 20/06/2000
The 2nd meeting of the CT Users Group was held in London on 20/06/2000. The programme is shown below.
Please note: information provided in the slides is not peer-reviewed, is for educational use only and is explicitly not to be used for sales or marketing purposes. Any of the authors can be contacted, via the CTUG if no contact information is provided in the slides, to discuss the contents.
| « Previous meeting | – | Meetings Index | – | Next meeting » |
|---|
Meeting Programme
10:00 Electron beam CT: A Physicist’s Perspective - I A Castellano Smith - Royal Marden Hospital, London
Electron beam CT (EBCT) was introduced in 1985 by Imatron Inc. The primary function of the scanner was and still is to produce CT images with minimal motion artefacts in order to optimise the visualisation of fast-moving anatomical features such as the heart and the lungs. In this presentation the design and operation of the scanner is reviewed, including design changes that have been incorporated into more recent models. Quality control is considered, with descriptions of appropriate dosimetric and image quality tests and testing frequency. Typical results are given and where appropriate compared with those obtained on other CT scanners; patient doses, described in terms of dose-length product, are also considered. The advantages and disadvantages of this technology as perceived by the clinician are identified. In conclusion the author gives a professional opinion on the strengths and weaknesses of the scanner performance, based on experience.
10:30 A comparison of the design features of multi-slice CT scanner models - Anne Hill, Maria Lewis, Sue Edyvean - ImPACT
Since the introduction of CT scanning into clinical practice in the early 1970s, the technology has matured greatly. The latest innovation, multi-slice CT scanning, allows the simultaneous imaging of four slices in a single rotation. This, together with reduced rotation times, results in decreases in examination time of up to a factor of seven, for a similar image quality as a single slice, 1-second scanner. A significant number of multi-slice systems are already in clinical use world-wide. In the UK four systems had been installed by November 1999, and it is anticipated that this figure will increase rapidly. Four CT manufacturers currently offer multi-slice scanners, based on three different design principles. Due principally to variations in the detector array design and rotation drive mechanism, the systems have varying capabilities with respect to the length imaged in a single rotation, range of slice widths available and rotation speed. The scanners also have differing specifications for reconstruction times and x-ray tube heat capacity, as well as for the standard image quality and dose parameters. The various multi-slice scanner designs are presented and their capabilities compared. Implications of the design principles on image quality and dose are discussed.
10:50 A Novel Approach to Patient Dosimetry in CT - D McRobbie - Hammersmith Hospitals NHS Trust
For the best part of a decade, patient dosimetry for CT within the UK has relied upon the Monte Carlo methodology developed by the NRPB. Whilst this method has been refined and updated to include new CT scanners and represent new features such as spiral scanning, its reliance on approximations of scanner geometry and a particular mathematical anatomical model are limiting factors.
In an examination of organ data derived from the NRPB model, we came across the interesting observation than the organ doses from scattered calculated from Monte Carlo conform to a simple exponential relationship to the position of the irradiated slice. This led to develop a simple empirical model of patient dose, by considering a primary beam area and a scattered dose area, where the scattering attenuation factors can be determined empirically.
Data were acquired on a Siemens Plus 4 scanner using a Rando anthropomorphic phantom and linear attenuation coefficients for scatter determined for 5 and 10 mm slice width in cranial-caudal and caudo-cranial directions for a reference slice through the abdomen. Additionally, we investigated the in slice variation in primary beam and the radial variation of the scatter.
From this data we are able to produce a simple 1d model for dose deposition. In a further development we consider the examination dose to be a summation of primary and scattered dose profiles. Moreover we can extend this model to incorporate movement of the irradiated slice in spiral scanning thorough a convolution of the static dose profile and the couch movement. Fourier Transform theory allows us to simply estimate doses from arbitrary scans from the inverse transform of the product of transforms of the static dose profile and a scan length function. The convolution approach can also be applied to sequential scans if an appropriate discrete transform is used. Pilot data using this method compared the NRPB model will be presented.
Ref. NRPB R249 Survey of CT Practice in the UK: Part 2 Dosimetric Aspects 1991 Kiremidjian H, McRobbie D.W. & McBride A (1998) “A new Fourier-based dosimetry method for CT” Institute of Physics and Engineering in Medicine, Annual Conference, Brighton.11:20 Development Of New Phantoms For The Physical Assessment of Multi-slice CT - S A Sassi - ImPACT
Purpose: Multi-Slice scanners offer a wide range of slice thicknesses and scan combinations, which should be catered for in developing phantoms for the assessment of these scanners. With nearly two decades of experience in developing physical phantoms for CT scanner assessment at ImPACT, we embarked on developing new phantoms suitable for the assessment of Multi-Slice scanners.
Methods: Quad-Slice scanners have up to four times more volume coverage per rotation than Single Slice scanners scanned at the same pitch. This needs to be reflected in the extra length of the phantoms required. They also offer a wider range of slice thicknesses between 0.5mm and 10mm requiring consistently high contrast for widely variable partial volume averaging. The new phantoms can be divided into two categories; water filled phantoms for noise and uniformity measurements, and inserts, which fit into existing CTDI phantoms for high contrast spatial resolution and slice sensitivity measurements. The high contrast spatial resolution insert is a 0.1mm diameter 140mm long tungsten wire cast in Perspex. The z-sensitivity insert is a 6mm diameter 0.05mm thick tungsten disk embedded in Perspex.
Results: The phantoms and the individual inserts provide the tools for the physical assessment of spiral and axial CT image quality including image noise, uniformity, high contrast spatial resolution and slice sensitivity. Preliminary evaluation data has been obtained from single slice GE CT/i scanner and a GE Lightspeed QX/i Multi-Slice scanner.
Conclusion: The new phantoms meet the increased demands of spiral Multi-Slice CT on phantom design, particularly z-axis length and uniformity.11:40 Can Multi-slice Helical CT Achieve A Better Noise-Dose Relationship Than Single Slice? - Julia Carden - ImPACT
Manufacturers claim that multi-slice CT scanners exhibit performance peaks at certain helical pitches, whereas in single-slice CT the image noise remains unchanged with pitch and the slice sensitivity profile steadily deteriorates with increasing pitch. The inference is that, by combining optimised pitch with a new form of helical interpolation algorithm, an improved relationship between noise and dose at a given effective slice thickness can be achieved in multi-slice scanning compared with single-slice scanning. Most of the evidence that has appeared in the literature is based on computer simulations. The aim of our study was to obtain experimental data for the relationship between noise, dose and slice profile at the available pitches on both a multi-slice and a single-slice scanner. We used standard ImPACT phantoms and protocols to measure noise, dose and slice profiles on a single-slice IGE HiSpeed CT/i scanner and on a multi-slice IGE LightSpeed QX/i scanner. A comparison of the results demonstrates the extent to which the manufacturer’s claims for multi-slice CT performance are borne out.
12:00 CT QA: The Radiotherapy Perspective - L Kadiri - Royal South Hants Hospital
In CT image based radiotherapy planning (RTP), data of patients scanned in the treatment position are used on a treatment planning system (TPS). The accuracy of RTP depends on existence of the usual CT QA for safety and diagnosis. Supplementary QA of the data transfer and the TPS reconstructed image is necessary. This may also include checks of the CT alignment mechanism and the patient positioning aids. In this paper, we discuss the methods and equipment for the additional QA using two examples. The first was a phantom designed for the QA of three dimensional RTP at Southampton. The second was a QA programme developed at the Suffolk Oncology Centre that satisfied the local RTP requirements without increasing the time for CT QA. The results were cited. It was concluded that these checks complement the usual QA of the CT and the treatment planning system.
12:15 A Comparison of Methods for Measuring the Limited Spatial Resolution of a CT Scanner - Bernhard Warr - ImPACT
The use of CT as a widespread diagnostic imaging tool necessitates the accurate assessment of a scanner’s capabilities, specifically spatial resolution. The test tool provided for the assessment of high contrast spatial resolution is usually a bar pattern for subjective visual analysis. The design of these bar phantoms varies between different manufacturers, and although suitable for QA procedures, it does not allow for reliable comparisons between different scanner models. Therefore, it is important that these subjective checks can be related to one another, as well as the more objective methods used by the ImPACT group. The aim of this project was to determine the correlation between techniques so that fair comparisons can be made when required. Measurements were conducted on an IGE HiSpeed CT/i scanner recently installed at St.George’s Hospital, London. We used a range of both objective and subjective methods. These included Fourier analysis of Edge Spread Function data, from our own phantom, to calculate MTF 10% and 50% values, the Droege method of analysis (related to the modulation of a bar pattern) on IGE’s own QA phantom and visual analysis on a ‘Catphan’. By upgrading our image analysis software we were able to test a new wire insert made to fit into one of ImPACT’s existing phantoms, which allows us another method of objectively measuring, via the Point Spread Function, MTF 10% and 50% values for, axial, helical and multi-slice scanning. Results are presented showing the correlation between the different methods.
13:50 Radiation Doses in CT Fluoroscopy - D McRobbie, AKP Lim, SJ Higgins, N J Davies, A Hancock, ME Roddie - Hammersmith Hospitals
CT fluoroscopy is a new real time imaging technique which allows greater control and accuracy for biopsy guidance using CT. Data will be presented from an initial 21 patients undergoing CT guided biopsy using a Siemens Plus 4 scanner with CARE Vision. Parameters to be reported include patient skin doses, operators' hand and eye doses and screening time. All doses were measured using TLDs. Comparisons will be made with other figures from the radiology literature. Comments on fluoroscopic technique will be made.
14:00 Stability of CT numbers over a long period of time - Jonathan Eatough - Addenbrookes Hospital
The IPEM and the European Study Group (ESG) give quite different quality criteria for stability of CT numbers over time. In East Anglia measurements have been made on a dozen scanners at approximately six monthly intervals using a phantom with a range of inserts. Measurements have been carried out over a period of up to 12 years using the same phantom. The results confirm the theoretical prediction that it is not appropriate to set a single criterion for stability across the full range of CT numbers. With modern equipment the CT number should remain within +/- 5 Hounsfield units or +/- 5% of the baseline value whichever is the greater. Meeting these criteria will assist in demonstrating that image quality is being maintained and will also ensure acceptable precision for radiotherapy treatment planning.
Round table discussions
14:15 Experience of kV measurements - Alexis Moore, Leeds General Infirmary
14:25 Is Your Tube Correctly Aligned? - Matthew Dunn, University Hospital Birmingham
14:50 The CEC Criteria and Application - Jennifer Poveda, Nottingham City Hospital
15:00 kV and HVL Measurements - Jonathon Eatough
15:20 Using NRPB SR250 for Spiral Scanning - Matthew Dunn, University Hospital Birmingham
15:30 Cone Beam Spiral CT (Video Presentation) - Ge Wang, PhD, Associate Prof. Department of Radiology, University of Iowa
