RPS 1213 - Evolution of CT: a key to its sustainability

Author Block: M. C. Daniotti¹, S. Trivellato², L. De Sanctis¹, V. Pisoni², J. Stancanello³, J. Mason³, R. Pellegrini³, S. Arcangeli², E. De Ponti²; ¹Milan/IT, ²Monza/IT, ³Stockholm/SE
Purpose: A new CBCT reconstruction algorithm based on poli-energetic quantitative (Polyquant) method empowered with a convolutional neural network scatter correction has been recently proposed. This study aimed to validate the use of the Polyquant CBCTs (pCBCTs) for image-guided radiotherapy (IGRT) and planning for locally advanced rectal cancer (LARC).
Methods or Background: Translational shifts obtained on all axes with the registration of pCBCTs to CT were compared to the clinical version of Elekta XVI CBCT ones and statistical significance was investigated with the t-test and ANOVA-test.
pCBCTs were calibrated with a population-based curve (pop-CC) elaborated coupling pCBCT gray levels to the CT relative electron density (RED) for ten pelvic patients. pop-CC was validated by comparing dose calculations on pCBCT and bulk density pseudo-CT using 1%/3mm local gamma-analysis.
Five LARC patients treated on Elekta VersaHD were selected. The RED difference between CT and first session-pCBCT were assessed on a voxel-to-voxel basis, on a contour basis, and on a dosimetric basis using 1%/3mm local gamma-analysis to compare pCBCT and pseudo-CT calculations.
Results or Findings: The translational differences between IGRT results of CBCT and pCBCT were always <1mm and not statistically significant.
The pop-CC pCBCT calibration resulted in dose calculations comparable with the pseudo-CT ones, with gamma passing rates > 95%.
For LARC patients, voxel-to-voxel and structure-based analysis showed no relevant RED discrepancies between pCBCT and CT. Residual RED differences resulted dosimetrically negligible compared with dose distributions calculated on pseudo-CT, with gamma passing rates > 95%.
Conclusion: The optimized pCBCTs were successfully RED calibrated and validated for IGRT and planning for LARC radiotherapy. The results suggest that pCBCTs could be exploited in the clinical workflow for adaptive radiotherapy in LARC patients,
Limitations: Further investigation for their extended use might still be necessary.
Funding for this study: No funding
Has your study been approved by an ethics committee? Not applicable
Ethics committee - additional information: Not applicable

Author Block: E. Verelst¹, G. Van Gompel¹, D. Crotty², H. Linder³, P. D. Deak⁴, J. De Mey¹, N. Buls¹; ¹Brussels/BE, ²Cork/IE, ³Stockholm/SE, ⁴Münsingen/CH
Purpose: Reducing motion-induced artefacts is an increasingly important aspect of ultra high-resolution (UHR) photon-counting CT (PCCT). Using a motion-controlled phantom, this study evaluates the benefit of using wider detector coverage in PCCT to reduce motion artefacts.
Methods or Background: A tissue-simulating cuboid phantom, 80-mm in length, was mounted onto a custom-built motion-controlled phantom. A 3-mm diameter stent (Superia, Nano-Therapeutics, India) was inserted into a commensurate hole inside the phantom. Programmed to simulate pulsatile motion, helical images were acquired during motion by a prototype Silicon-based PCCT (Si-PCCT) using 40- and 80-mm detector coverage, representing table speeds of 128.6-mm/s and 257.1-mm/s, respectively. Motion-induced image artefacts were evaluated by measuring the volumetric error relative to the nominal phantom volume. A 5-point Likert rating evaluated stent appearance for both detector coverages against a reference static image (1-no similarity, 5-similar). Differences were assessed using a paired sample t-test and Wilcoxon signed-rank test, respectively. P-values < 0.05 indicated statistical significance.
Results or Findings: Relative to the nominal phantom volume (1571-mm3), the 80-mm detector coverage statistically significantly reduced volumetric error (mean 81-mm3, SD 18-mm3) compared to the error generated with 40-mm coverage (254-mm3, 48-mm3), p=0.006. Stent appearance under wider detector coverage was likewise assessed to better match the reference image, with average Likert scores for 40 mm and 80mm of 1 [1–1.5] and 4 [3.5–3.75], respectively, p=0.043.
Conclusion: In high-resolution PCCT imaging, to combat motion-induced artefacts, it is important to combine wide-detector UHR CT acquisitions with higher table speeds. This study demonstrates the potential of using a prototype wide-coverage Si-PCCT system with fast tables speeds to reduce such artefacts. While this study specifically evaluated pulsatile motion, an 80-mm detector coverage may reduce additional body motion-induced artefacts, such as peristalsis.
Limitations: This is an ex-vivo phantom study.
Funding for this study: Flemish Research Foundation (FWO), personal grant, nr: 1SH1Z24N.
Has your study been approved by an ethics committee? Not applicable
Ethics committee - additional information: Ethical approval was not required for this study

Author Block: K. Mei¹, L. Roshkovan¹, S. Sharma², S. Ross², J. Woo¹, S. S. Halliburton³, L. P. Liu¹, R. Thompson³, P. Noël¹; ¹Philadelphia, PA/US, ²Vernon Hills, IL/US, ³Mayfield Village, OH/US
Purpose: To develop a 3D-printed, patient-specific brain phantom for assessing performance of non-contrast CT head imaging.
Methods or Background: Unenhanced T1 MRI scan (best available gold standard) of a healthy brain (21y/o,F) was retrospectively collected and converted into CT Hounsfield unit image to generate a realistic phantom using PixelPrint technique. The brain phantom was created as a 30 mm thickness section including both left and right hemispheres (approximately 157 x 120 mm) at 1:1 scale. Additionally, a separate skull phantom was printed from the same patient images using calcium-doped filament.
The brain phantom, with surrounding skull phantom, was scanned with CT at 120 kVp and 24.4 mGy. Images were reconstructed with and without iterative denoising at 0.5 mm pixel spacing. Attenuation values were measured in gray and white matter.
Results or Findings: Gray and white matter were clearly distinguishable in an appropriate CT examination window, with and without denoising. Line profile was plotted along the center of the phantom. Realistic CT values of approximately 45 HU for gray matter and 25 HU for white matter were observed. Maximum density observed in the skull reached approximately 750 HU, which was limited by the density of the 3D-printing filament used. Image noise, estimated by standard deviation, ranged between 3 and 4.5 HU across both sets of denoised images.
Conclusion: The PixelPrint 3D-printed brain phantom successfully depicts realistic tissue attenuationfor white and gray matter, demonstrating potential as a valuable tool for evaluating CT head imaging performance.
Limitations: This study converts MR images to CT numbers and simulates non-contrast CT scans.
Funding for this study: This work was partly supported by Canon Medical Systems Corporation (Otawara, Japan).
Has your study been approved by an ethics committee? Yes
Ethics committee - additional information: University of Pennsylvania

Author Block: A. J. Gutwinska¹, D. Rosario², A. Pang², C. A. Hernandez-Fajardo¹, R. Coulibaly¹, A. Robert¹, S. Rit¹, D. P. Cormode², S. A. Si-Mohamed³; ¹Lyon/FR, ²Philadelphia, PA/US, ³Bron/FR
Purpose: To evaluate the image quality of color K-edge imaging for contrast agents based on 8 different elements using a clinical prototype spectral photon-counting CT (SPCCT).
Methods or Background: A SPCCT with a field-of-view of 500mm was used (Philips; Israel). An anthropomorphic thoracic phantom (QRM GmbH) with twelve 1.5mL K-edge solutions (gadolinium-Gd, holmium-Ho, ytterbium-Yb, hafnium-Hf, tantalum-Ta, tungsten-W, gold-Au, bismuth-Bi) ranging from 0 to 2mg/mL was scanned at 120kVp and 50/75/150mAs. Five acquisitions per agent were performed using dedicated energy thresholds. Conventional images in Hounsfield units and color K-edge images in mg/mL were reconstructed with isotropic voxels of 0.7mm3. Noise, mean relative error (MRE) between prepared and measured concentrations, signal-to-noise ratio (SNR) were measured on color K-edge images. Contrast-to-noise ratio (CNR) on conventional and color K-edge images were measured and compared.
Results or Findings: Mean noise ranged from 0.04 to 0.13mg/mL among all samples with a lowest value for Gd at 75mAs (0.04±0.01 mg/mL) and highest for Ta at 50mAs (0.13±0.03mg/mL). Overall MRE was 29.7% with higher accuracy for Yb (e.g., 7.0% at 150mAs), and lower for Ta (55.7% at 150mAs). SNR increased as function of concentrations with a factor per mg of 16.7, 14.6, 10.9, 9.5, 9.3, 9.3, 8.8, 4.7 for Gd, Yb, Ho, Hf, Au, W, Ta and Bi, respectively, at 75mAs. CNR in color K-edge images increased as function of concentrations, and were higher in comparison to CNR in conventional images (e.g., 1075%, 957%, 677%, 658%, 601%, 590%, 536%, 411%, for Gd, Au, W, Ta, Yb, Ho, Bi and Hf, respectively, at 75mAs).
Conclusion: Image quality of color K-edge imaging in an anthropomorphic thoracic phantom demonstrated high performances for 8 color K-edge agents using SPCCT whilst outperforming sensitivity in comparison to conventional imaging.
Limitations: Phantom study.
Funding for this study: The ERC starting Grant "KOLOR SPCCT Imaging" (N°101118079).
Has your study been approved by an ethics committee? Not applicable
Ethics committee - additional information: Not applicable.

Author Block: R. Coulibaly¹, A. Robert¹, A. Houmeau¹, M. N. Antonuccio², P. C. Douek¹, S. Rit¹, J. Greffier³, S. A. Si-Mohamed¹; ¹Lyon/FR, ²Paris/FR, ³Nimes/FR
Purpose: To evaluate the image quality of color K-edge imaging with a gadolinium agent using spectral photon-counting CT (SPCCT) in a phantom with a mixture of contrast agents.
Methods or Background: A clinical prototype SPCCT system (FOV 500mm, Philips; Israel) was used to scan custom-made cylindrical phantom of 27cm (Color iQCT). Three inserts of the phantom were filled up with agents as follows: iodine only, mixture of iodine and gadolinium, gadolinium only. Two configurations were considered, one with 0.5mg/mL of each contrast agent and another with 2mg/mL. For each configuration, two series of nine helical scans (120kVp) were acquired at 75mAs and 150mAs. Spectral K-edge images of gadolinium were obtained by doing material decomposition using 3 basis (water/iodine/gadolinium), using an iterative reconstruction algorithm at 3 levels (iDose 0, 6, 11) were compared between inserts with gadolinium using iQMetrix-CT software.
Results or Findings: Despite the presence of iodine, color K-edge imaging enable specific differentiation of the gadolinium, showing a concentration difference of 1.1% between inserts of gadolinium only and mixture (150mAs, i11, 2mg/mL). NPS peak was observed at the same spatial frequency in all configurations (i.e., 0.053±0.018 mm⁻¹), whereas the noise magnitude decreased when the dose and the iDose4 levels increased (-19.65±0.01% between iDose0 and iDose11). TTF values at 50% (f50) were similar between inserts with gadolinium only and the mixture, regardless of the dose or iDose levels (e.g., 0.213±0.090 mm⁻¹ vs 0.237±0.061 mm⁻¹, at 150mAs, i11, 0.5mg/mL). f50 values were improved with increasing dose, iDose levels, and with higher concentrations.
Conclusion: Color K-edge imaging of a gadolinium contrast agent using SPCCT demonstrated high spatial resolution and low noise magnitude at low concentrations, even in the presence of iodine.
Limitations: Phantom study.
Funding for this study: ERC starting Grant "KOLOR SPCCT Imaging" (N°101118079).
Has your study been approved by an ethics committee? Not applicable
Ethics committee - additional information: No ethics was required for this phantom study.

Author Block: G. Havariyoun; London/UK
Purpose: Several surgical procedures benefit from the ability to image resected tissue samples in real time, e.g. to ensure no margin involvement. Micro-CT or tomosynthesis have great potential, but suffer from limited soft-tissue sensitivity of X-rays. X-ray phase contrast imaging (XPCI) provides soft tissue sensitivity and increased contrast through exploitation of phase effects. This work is presented on behalf of the UCL AXIm team.
Methods or Background: XPCI was initially restricted to specialized facilities such as synchrotrons. Our group has developed a method that has enabled creation of a pre-commercial prototype compatible with surgical and radiology workflows. This has been used to image >100 breast tissue samples from breast conserving surgery both in vitro and in real time. Images were compared to standard specimen radiography and histopathology. The system also allows higher resolution (~10 micrometre) imaging in slower scans for e.g. digital histology.
Results or Findings: System optimization (which also included size reduction) led to clinically acceptable scan times, which were verified by trialling the system in a real intra-operative context. XPCI imaging resulted to sensitivity and specificity values of 83% (95% CI 69-92%) and 83% (95% CI 70-92%), respectively. Standard specimen radiography resulted to sensitivity and specificity values of 32% (95% CI 20-49%) and 86% (95% CI 73-93%), respectively.
Conclusion: XPCI has a specificity comparable to standard specimen radiography but a significantly higher sensitivity. This would lead to significant reduction in re-excision rates and in turn a reduction in patient stress, surgical times, healthcare costs and improved cosmetic outcomes.
Limitations: Comparisons with standard specimen radiography were made as this is the most commonly used tool in the clinical setting. Comparison with more advanced techniques will be made in the future.
Funding for this study: This work is funded by the Wellcome Trust (Grant 200137/Z/15/Z). Alessandro Olivo (AXIm lead) is funded by the Royal Academy of Engineering under their “Chairs in Emerging Technologies” scheme (CiET1819/2/78).
Has your study been approved by an ethics committee? Yes
Ethics committee - additional information: The Breast Cancer Now Tissue Bank (Approval No. 15/EE/0192) provided the ethically approved samples, the authors thank the patients who have generously consented to donate their tissues which have been utilised in this work.

Author Block: J. Vignero, B. Miseur, J. Binst, H. Bosmans; Leuven/BE
Purpose: A radiologist had raised concerns about excessive noise in CT abdomen images of scanner A, while another scanner (B) of the same model, using similar protocol settings, produced images of acceptable quality. This discrepancy led to a study aimed at harmonizing protocol settings to ensure consistent image quality across all patient sizes, moving beyond standard dose equalization methods. Currently, protocol adjustments—based on dose level, scan task, and tube current modulation strength—are often made intuitively.
Methods or Background: Clinical CT scan data, collected through a dose monitoring platform (DOSE, Qaelum), was analysed using water equivalent diameter (WED), global noise level (GNL), kVp and CTDI. Three phantoms with ellipsoid cross-sections (WED: 22, 33 and 43cm) and iodine/calcium inserts were scanned to map protocol settings to scan parameters (kVp, mAs, CTDI) and GNL for each scanner and phantom. The reconstruction settings were kept fixed.
Results or Findings: Initially, scanner A and B produced similar radiation doses for the same WED groups, but scanner A had on average a 13% higher GNL, with a maximum of 18% difference for the smallest WED group. The radiologist could define GNL upper limits for each kVp. Using the phantom scan maps, protocol settings were chosen to achieve the desired image quality. Post-optimization, 87% of scans met the GNL criteria, compared to 70% before. To achieve this, radiation doses were increased with 17%.
Conclusion: Identical scanner models and protocol settings do not guarantee consistent image quality. Quality measures should be included in optimization efforts. We present a procedure using new phantoms and dedicated metrics to improve protocol harmonization.
Limitations: The pipeline has only be verified on two scanners.
Funding for this study: In part funded by the i-Violin project that is co-funded under the EU4Health Programme 2021-2027, grant agreement no. 101056832.
Has your study been approved by an ethics committee? Not applicable
Ethics committee - additional information: Retrospective, technical study

Author Block: O. Sandvold¹, A. Perkins², H. Daerr³, T. Koehler³, R. Proksa¹, R. Manjeshwar², P. Noël¹; ¹Philadelphia, PA/US, ²Cleveland, OH/US, ³Hamburg/DE
Purpose: To investigate the quantitative effects of varying the ratio of high and low kVp tube voltage durations on spectral CT results with a rapid kVp-switching X-ray tube.
Methods or Background: Rapid kVp-switching offers excellent spectral separation, but experimental research on acquisition parameters is limited. This study addresses the gap.
A rapid kVp-switching X-ray tube (Philips Healthcare) on a spectral CT bench system was operated at 500 mAs alternating between 140 and 80 kVp. The total integration period (IP) containing high and low kVp IPs was 1 ms. We varied the ratio of 140 kVp duration to total IP time from 0.15 to 0.85 and correspondingly adjusted the low kVp IP time. A 3D-printed plastic phantom (20 cm diameter) was rotated at 1 Hz and contained four tissue-mimicking inserts: iodine (2.0, 5.0 mg/ml), iodine 4.0 mg/ml + human equivalent (HE) blood, and HE blood (Sun Nuclear). We performed reconstruction and two-material decomposition without applying denoising. The noise and contrast-to-noise ratio (CNR) of the known concentration inserts were measured in photoelectric material images. Dose was estimated from the reference detector.
Results or Findings: Dose comparisons showed 0.15, 0.25, 0.33, and 0.67 ratio scans used 47%, 51%, 63%, and 91% of the highest dose scan (0.85 ratio). Measured noise values appeared to follow a quadratic trend as the ratio increased, with 0.33 containing the lowest average noise. The 0.33 ratio image dose normalized CNR was approximately 1.24x, 1.14x, 1.11x, and 1.68x greater than the dose normalized CNR in 0.15, 0.25, 0.67, and 0.85 ratio images respectively.
Conclusion: Dose normalized CNR depends strongly on the ratio between high and low kV durations. This ratio should be considered for optimizing the spectral acquisition.
Limitations: None
Funding for this study: None
Has your study been approved by an ethics committee? Not applicable
Ethics committee - additional information: Not applicable