RPS 2204 - New techniques for pulmonary imaging

Author Block: F. T. Gassert, R. Burkhardt, T. Gora, D. Pfeiffer, A. Fingerle, A. Sauter, M. Makowski, J. Wilkens, F. Pfeiffer; Munich/DE
Purpose or Learning Objective: The aim of this study was to show the benefit of dark-field CT imaging for the detection of radiation-induced lung damage in early stages.
Methods or Background: We compared attenuation based CT imaging to dark-field (DF) CT imaging in a murine model of radiation-induced lung damage in the right lung (n=6) and a control group (n=6). Animals were scanned before irradiation and 12 weeks, 16 weeks, 20 weeks and 24 weeks thereafter. Three radiologists assessed the images twice for the presence of lung damage and rated their confidence on a scale from 1 to 5. The inter-rater- and intra-rater-reliability was determined and rated with Cohen’s 𝜅 respectively Fleiss’ 𝜅. For the quantitative analysis the ratio of the mean pixel value of the right and left lung (⟨𝑅⟩=mright/mleft) was calculated. Results of the irradiated group were compared with the respective control group by using a t-test.
Results or Findings: The sensitivity of DF CT for radiation-induced lung damage in the reader study was significantly higher at 12 weeks (Att: 36.7%, DF: 53.3%, p=0.023) and at 16 weeks (Att: 50.0%, DF: 91.7%, p<0.001). The overall confidence of the readers was significantly higher when reading DF images (Att: 3.48, DF: 4.77, p<0.001). Both the average intra-rater-reliability (Att: 𝜅=0.82 , DF: 𝜅=0.91) and the inter-rater-reliability (Att: 𝜅=0.66, DF: 𝜅=0.75) were higher for DF imaging. For attenuation based imaging the difference of the ratio ⟨𝑅⟩ between the control group and the irradiated group became significant after 20 weeks (p=0.011), while for DF imaging it was already highly significant after 16 weeks (p=0.003).
Conclusion: This small animal study demonstrates that dark-field CT imaging allows for the detection of radiation-induced lung damage in early stages and, in that respect, is superior to conventional CT.
Limitations: This study is limited by its small cohort only.
Ethics committee approval: The ethics committee approval was obtained.
Funding for this study: Not applicable.

Author Block: S. A. Si-Mohamed¹, S. Boccalini¹, P-A. Rodesch¹, R. A. K. Dessouky², P. Coulon³, E. Lahoud⁴, M. Villien¹, L. Boussel¹, P. Douek¹; ¹Lyon/FR, ²Zagazig/EG, ³Dourdan/FR, ⁴Haifa/IL
Purpose or Learning Objective: To characterise the technical capabilities of a clinical spectral photon-counting CT (SPCCT) and to evaluate its feasibility on a first human volunteer for high-resolution lung imaging.
Methods or Background: Measurement of a modulation transfer function (MTF) and acquisition of a line pairs phantom were performed. An anthropomorphic lung nodules phantom was scanned under standard, low, and ultra-low radiation doses. A human volunteer underwent scans, at 120 kVp standard (62 mAs) and low (11 mAs) dose. High-resolution (HR) images were reconstructed with 1024 matrix, 300mm FOV and 0.25mm slice thickness. Lung structures conspicuity and sharpness, image noise and overall image quality were independently analysed by three radiologists and compared to a previous scan (120kVp, 10mAs).
Results or Findings: 10 % MTF was measured at 22.3lp/cm with a cut-off at 31lp/cm. Up to 28lp/cm were depicted. While mixed and solid nodules were depicted on standard and low dose images with FBP, ultra-low dose imaging necessitated the use of iDose and 1mm slice thickness to allow visualisation of the ground-glass component. In a human volunteer, standard dose SPCCT images were of greater overall image quality and lung structures conspicuity and sharpness than conventional CT images and comparable image noise. Low dose SPCCT images were of greater or similar conspicuity and sharpness of lung structures, of equivalent overall image quality, of lower but acceptable image noise despite a flux reduction of 89%.
Conclusion: An SPCCT prototype demonstrated high-resolution technical capabilities and high image quality on a human volunteer for lung imaging.
Limitations: Exhaustive evaluation of image quality was out of the scope of the present study opening the way for furthermore investigations.
Ethics committee approval: IRB approved the study.
Funding for this study: This study was funded by the European Union Horizon 2020 grant No 668142.

Author Block: L. Jungblut, T. D. J. Sartoretti, D. Kronenberg, V. Mergen, A. Euler, H. Alkadhi, T. Frauenfelder, K. Martini; Zurich/CH
Purpose or Learning Objective: The purpose of this study was to evaluate the accuracy of emphysema quantification as performed on post-processed virtual non-contrast images derived from photon-counting detector computed tomography (PCD-CT).
Methods or Background: Sixty-five patients who underwent a three-phase chest CT on a first-generation, clinical dual-source PCD-CT were retrospectively included. Scans were performed in the multi-energy (QuantumPlus) mode at 120kV with a weight-adjusted intravenous contrast agent. Virtual non-contrast images (VNC) were post-processed from the venous as well as from the arterial phase. Images were assessed quantitatively (global noise index (GNI)) and qualitatively by independent readers (overall image quality, emphysema assessment, delineation of small structures). Emphysema quantification (with a threshold of -950 HU) was performed for non-contrast images, contrast-enhanced images (arterial and venous phase) and the post-processed virtual non-contrast images (generated from both; the atrial and the venous phase) by commercially available software. Non-contrast images served as a reference standard for emphysema quantification.
Results or Findings: Virtual non-contrast images post-processed from the arterial phase (p=0,409) as well as from the venous phase (p=0.093) showed no significant difference in emphysema quantification relative to true non-contrast images while there was a highly significant difference compared for the contrast-enhanced scans (arterial and venous; p<0.001). GNI showed no significant difference between the virtual non-contrast image from the arterial and venous phase and the true non-contrast image. The score of subjective assessment was highest for the true non-contrast image (p<0.001) while there was no significant difference between both virtual non-contrast reconstructions.
Conclusion: Computer-aided emphysema quantification with PCD-CT is feasible for virtual non-contrast-enhanced images post-processed from the venous as well as the arterial phase.
Limitations: This study has been performed as a single-centre study.
Ethics committee approval: The ethics committee approved this study.
Funding for this study: No funding was received for this study.

Author Block: A. Bodenberger, P. Konietzke, O. Weinheimer, H-U. Kauczor, W. Stiller, T. F. Weber, T. D. Do, M. O. Wielpütz; Heidelberg/DE
Purpose or Learning Objective: Chronic-obstructive airway diseases show progressive remodelling of airway dimensions. Inflammatory activity may be visualised by contrast enhancement in computed tomography (CT) examinations, which has not been systematically studied to date. Spectral CT offers the possibility to study contrast enhancement without the need for multiphase acquisitions.
Methods or Background: 234 lung healthy patients underwent dual-layer spectral CT (Philips iQon) with four retrospective groups: non-enhanced (NE), pulmonary venous (PV), pulmonary arterial (PA) and systemic arterial (SA) standardised contrast phase. Ten virtual monoenergetic series were reconstructed at 40-160 keV. Fully automatic segmentations were carried out using validated in-house software. Attenuation of the airway wall was assessed in Hounsfield Units (HU) for airway generations G2, G3, G4 and combined G5-10. The slope of the spectral attenuation was calculated by λ=(HU40keVHU100keV)/60.
Results or Findings: Slopes were significantly different for all airway generations between NE (λNE,G5-10=0.3HU/keV) vs. contrast-enhanced acquisitions PV (λPV,G5-10=0.67HU/keV), PA (λPA,G5-10=1.83HU/keV) and SA (λSA,G5-10=1.97HU/keV) (λNE,G5-10-λPV,G5-10 p=0.006; λNE,G5-10-λPA,G5-10 p<0.001; λNE,G5-10-λSA,G5-10 p<0.001). Additionally, the slope differs between PV and PA phase (pG5-10<0.001) but does not vary between PA and SA phase (pG5-10>0.999). Wall thickness (WT) for generation G5-10 did not change significantly between groups at any keV level (40 keV: WTNE=1.14 mm, WTPV=1.22 mm, WTPA=1.33 mm, WTSA=1.22 mm, p=0.054).
Conclusion: Spectral CT may quantify airway wall attenuation with a single acquisition by determining the slope of spectral enhancement, and may separate arterial and venous enhancement. Further studies are warranted to analyse spectral CT for inflammatory airway diseases.
Limitations: The spectral slope is only assessed in airway-healthy patients.
Ethics committee approval: This retrospective study was approved by the institutional ethics committee (S-924/2019).
Funding for this study: This study was supported by grants from the Bundesministerium für Bildung und Forschung (BMBF) to the German Center for Lung Research (DZL) (82DZL004A, 82DZL004A2).

Author Block: Y. Ohno¹, N. Akino², Y. Ito², H. KImata², K. Fujii², Y. Fujisawa², K. Murayama¹, Y. Kataoka¹, H. Toyama¹; ¹Toyoake/JP, ²Otawara/JP
Purpose or Learning Objective: To compare radiation dose reduction capability of ultra-high resolution CT (UHR-CT) and area-detector CT (ADCT) for lung density evaluation among hybrid-type and model-based iterative reconstruction (IR) and developed deep learning reconstruction (DLR) at Quantitative Imaging Biomarkers Alliance (QIBA) recommended phantom study.
Methods or Background: QIBA recommended phantom was scanned by UHR-CT with normal resolution (NR: 0.5mm×80 rows/896 channels), high resolution (HR: 0.5mm×80 rows/1792 channels) and super-high resolution (SHR: 0.25mm×160 rows/1792 channels) and ADCT (0.5mm×80 rows/896 channels) at 400mA, 230mA, 100mA, 50mA, 20mA and 6mA in five times. Then, all CT data were reconstructed as 0.5mm and 1mm section thicknesses by each reconstruction. Then, CT values of all density forms and CT value within lung density form were determined by ROI measurements five times. Pierson’s correlation was analysed between measured CT density and each form density on all CT protocols. To compare the capability for radiation dose reduction on each CT data with ADCT obtained by 400mA and reconstructed with hybrid-type IR (i.e. standard protocol), ΔCT of each protocol was compared with that of the standard protocol by paired t-test.
Results or Findings: There was a significant and excellent correlation with standard reference on each protocol (0.990.05).
Conclusion: For lung density assessment, hybrid-type and model-based IR, as well as DLR on UHR-CT and ADCT, can reduce 95% radiation dose with keeping accuracy as compared with standard CT protocol and QIBA profile.
Limitations: Not applicable.
Ethics committee approval: This study was a phantom study and no need for IRB approval.
Funding for this study: This study was financially and technically supported by Canon Medical Systems Corporation.

Author Block: T. Urban¹, A. Sauter², M. Frank¹, K. Willer¹, T. Koehler³, F. T. Gassert², M. Makowski², D. Pfeiffer², F. Pfeiffer¹; ¹Garching/DE, ²Munich/DE, ³Hamburg/DE
Purpose or Learning Objective: Dark-field radiography can provide information on the condition of the lungs’ alveolar structure. It has recently been translated from the lab to the clinical stage. Here, we evaluate its performance for the detection of emphysema, and compare its diagnostic value with conventional radiography.
Methods or Background: We included 91 patients after a medically indicated CT scan, either without any lung impairment or with varying stages of emphysema. As a reference standard, visual scores based on the Fleischner scale for emphysema severity (absent, trace, mild, moderate, confluent, advanced destructive emphysema) were assigned to all CTs by three radiologists. For dark-field chest radiography, we employed a clinical prototype, which is capable of acquiring dark-field and attenuation-based radiographs simultaneously at a dose in the range of 0.035mSv. Both modalities, displayed individually and simultaneously, were rated for presence and severity of emphysema (no, mild, moderate, severe) by three radiologists. Statistical analysis included receiver-operator-characteristics and comparison of adjacent groups using two-sided Mann-Whitney-U-tests with a significance level of 0.05.
Results or Findings: The dark-field images showed a decrease in signal strength with emphysema severity. Compared to conventional images (AUC=0.73), readers were better able to identify mild emphysema when reading dark-field images (AUC=0.86). While the differentiation between trace and mild emphysema was not possible reading conventional radiographs, readers could differentiate between these stages based on dark-field radiographs.
Conclusion: Dark-field radiography increases the diagnostic value of attenuation-based radiography for the identification and staging of emphysema, especially in the early stages.
Limitations: There is only a limited number of participants. Emphysema was the only lung pathology under investigation.
Ethics committee approval: Approval of the Institutional Review Board was obtained prior to this study (IRB reference number 166/20S). All participants gave written informed consent.
Funding for this study: This study was funded by the European Research Council, Philips Medical Systems DMC GmbH, Karlsruhe Nano Micro Facility.