23rd CT users group meeting: 21/10/2021
The 23rd CT Users Group meeting was held online on 21/10/2021. The programme is shown below with links to pdf version of some of the talks.
Please note: information provided in the slides is not peer-reviewed, is for educational use only and is explicitly not to be used for sales or marketing purposes.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
13:30 Welcome/introduction - Katie Howard
Session 1 - Quality Assurance
13:35 Preliminary investigation of three Spectral CT Techniques and creation of a Quality Control programme - Laurence King, Royal United Hospitals Bath NHS Foundation Trust
A Multi Energy CT phantom (Sun Nuclear Corporation) was used to partially characterise the performance of three Spectral CT techniques on Siemens scanners at the Royal United Hospital, Bath. Phantom acquisitions were made using Siemens Dual Source, Dual Spiral and Twinbeam split filter techniques on the Somatom Drive and Somatom Edge+ model scanners. MonoE+ HU values, iodine quantification and rho/Z analysis was performed using the Syngo Via platform and compared across the techniques. This work is intended to inform the creation of a routine Spectral CT quality control programme.
In this presentation, the main manufacturers’ implementations of Spectral CT acquisition will be summarised, results of this work will be presented, and issues and relevant considerations around the design of a Spectral CT quality control programme will be discussed.13:55 Commissioning an AIRO Mobile CT Scanner - Adrian Porter , Laura Tonks, Dean Taylor, Barking, Havering and Redbridge University Hospitals Trust
The process involved with commissioning this system will be presented together with image quality and dosimetry results.
14:15 Evidence based QC for CT Scanners - an update from the IPEM working party - Gareth Iball, Mandy Price, IPEM Evidence Based QA Working Party
This working party was formed to assess the efficacy of x-ray quality assurance tests recommended in IPEM report 91, including CT scanners. For CT data, physics departments were contacted, primarily through the Medical-Physics-Engineering mailbase, to fill in a spreadsheet about which tests are performed at their centre and to give detailed results for each test performed. The assessment of efficacy considers the frequency with which a test result fails the remedial level from IPEM report 91. The third quartile of those results exceeding the remedial level or threshold is used to estimate the severity of such a breach in terms of potential impact on patient dose and image quality. A risk assessment approach is then used to recommend whether the test should be included in an on-going test schedule and if the current tolerances are appropriate.
Data was analysed from 485 surveys on 140 unique CT scanners throughout the UK. Across all tests, the rate with which the remedial level was exceeded varied from 0–6.4%, for tests with sufficient data for meaningful analysis. Severity ratings ranging from little or none to major degradation to image quality or significant increase on population dose were calculated.
Analysis of this data has led to some discussion around tolerances for image quality and dose tests and we would welcome feedback from the CT users on our ideas and methods.Session 2 - Protocol Optimisation & Dose Audit
14:50 Development of an optimal protocol for low dose CT imaging of metallic pelvic fixation devices - Ruby Callister and Jennifer Robinson, St George’s Hospital NHS Foundation Trust
Aim: Trauma and orthopaedics requested a new CT protocol to look at the position of metallic pelvic fixation devices shortly after surgery. This protocol needed to be optimised for visualisation of metal work and does not require the same image quality as a full diagnostic pelvic CT scan. Therefore, in theory, these exams can be performed at a much lower dose. This project investigated the optimal scan parameters for this clinical application.
Method: New protocols were established for two GE scanners, a Revolution and a Revolution GSI. Optimal parameters were determined based on phantom testing using the pelvis of a RANDO phantom. Metallic fixation devices were taped to the surface of the phantom and inserted between slices to mimic patient scans. Testing was performed to look at the effects of: pitch, kV, metal artefact reduction software, noise index and reconstruction settings. Image quality was assessed visually on volume rendered images and using noise measurements in a central uniform region on the phantom. Once the noise index was decided on the Revolution scanner, this was then translated to the Revolution GSI by matching levels of image noise.
Results: It was found that the use of high kV reduced the appearance of metal artefacts significantly and improved visualisation of the metallic devices. Using a pitch of 0.516 was found to be approximately 20% more dose efficient than a pitch of 0.981 and deemed appropriate as increased scan time is not a concern in this patient group. This pitch also reduced the appearance of metallic artefacts. Volume rendered images were found to be most useful for clinicians and their appearance was optimised by creating them from a soft tissue reconstruction with overlapping slices and no metal artefact reduction applied. This new protocol achieved adequate image quality whilst reducing patient dose by a factor of four compared to the routine diagnostic pelvis protocol. Feedback from clinicians regarding patient exam image quality has been positive so far and will be used to further optimise.
Conclusion: A low pitch, high kV, low dose protocol has been created which is suitable for post-operative evaluation of metallic fixation devices. The caveat must be made that it may not be suitable to diagnose soft tissue pathologies.15:10 CT Dose Audit - interpreting results with the aid of patient size evaluation* - Mike Holubinka, C Griffiths, R Mioduszewski , A Davis, Portsmouth Hospitals University NHS Trust and University of Portsmouth
Patient dose audits are increasingly undertaken remote to the imaging department utilising computer extraction of dose and all other information directly from DICOM headers and Radiation Structured Dose Reports (RSDR). Generally, and consistent with the earliest data collections, patient body size metrics are incomplete or missing entirely, there being no easy way to gather the data with the assistance of busy imaging staff. The exceptions to this may be paediatric imaging where entry of patient weight has been made a mandatory field.
Judging the data sample dose audit median against NDRLs and LDRLs is complicated by imprecise knowledge of the body size metrics of National and Local audit sample medians. Correct decisions in the optimisation process require knowledge of body size metrics if changes to protocols are to be made in pursuit of dose reduction or standardisation. Body sizes of Abdomen-Pelvis examinations were initially determined from axial images using automated extraction and calculation using ImageJ of Water Equivalent Diameter, Effective Diameter, cross-section area, AP and Lat dimensions. These were correlated with metrics for males and females of Normal Weight, Over-Weight, Obese level I, Obese level II and Morbidly Obese mathematical phantoms from the VirtualDose phantom sets. After matching audit sample sub-sets to the relevant phantom size, the weight adjusted CTDIvol and DLP values are calculated. For three of four scanners, the adjustment exceeded 30%, but less than 10% for the fourth. Additionally, the relationship of CTDIvol and DLP with patient size was clearly demonstrated, and also the relative change of indicative Effective Doses with body size, CTDIvol and DLP. Effective Diameter (EDia) was found to be a reliable body size metric at a consistent anatomical position, just superior to the iliac crest.
Extraction from axial images was both time consuming and data intensive owing to the number of images per study. As many CT systems utilise the AP and Lat topograms in scan parameter planning, it seemed logical to attempt to use these images in body size assessment. A modified Image J routine was employed and the Average Effective Diameter determined. Agreement with the axial EDia method was found to be good, and sufficient for dose audit and optimisation purposes. Automated send of all patient topograms to a local PC to cross-match with patient RDSR dose data is currently underway.
This would enable two approaches to data analysis. A) Automatically identify those patients of Normal Weight body metrics to select which dose data to include in the audit. B) Categorise all patient body sizes and construct CTDIvol/DLP vs body size curves for the relevant protocol and CT scanner. Neither approach negates whole sample dose median determination.15:30 AGM and officer election
Session3 - Discussion and updates
15:40 Frequent fliers in CT - Matt Dunn
15:50 CT Ionisation chamber calibration issues - David Platten
16:00 Verifying suitablility of new paediatric protocols before clinical use - Anne Hill
16:10 Image quality assessment for IR / deep learning reconstruction algorithms - Laurence King
16:20 Meeting Close
