PMC:7473770 / 43852-54017 JSON TXT

Annnotations TAB JSON ListView MergeView

LitCovid-PD-FMA-UBERON

LitCovid-PD-UBERON

LitCovid-PD-MONDO

{"project":"LitCovid-PD-MONDO","denotations":[{"id":"T59","span":{"begin":3546,"end":3554},"obj":"Disease"},{"id":"T60","span":{"begin":3622,"end":3633},"obj":"Disease"},{"id":"T61","span":{"begin":4011,"end":4019},"obj":"Disease"},{"id":"T62","span":{"begin":4434,"end":4442},"obj":"Disease"},{"id":"T63","span":{"begin":9667,"end":9669},"obj":"Disease"}],"attributes":[{"id":"A59","pred":"mondo_id","subj":"T59","obj":"http://purl.obolibrary.org/obo/MONDO_0000831"},{"id":"A60","pred":"mondo_id","subj":"T60","obj":"http://purl.obolibrary.org/obo/MONDO_0006654"},{"id":"A61","pred":"mondo_id","subj":"T61","obj":"http://purl.obolibrary.org/obo/MONDO_0000831"},{"id":"A62","pred":"mondo_id","subj":"T62","obj":"http://purl.obolibrary.org/obo/MONDO_0000831"},{"id":"A63","pred":"mondo_id","subj":"T63","obj":"http://purl.obolibrary.org/obo/MONDO_0019035"}],"text":"Funding Information\nThis paper was supported by the following grants:\nhttp://dx.doi.org/10.13039/501100002347Bundesministerium für Bildung und Forschung 05K19MG2 to Tim Salditt.\nhttp://dx.doi.org/10.13039/100010663H2020 European Research Council 771883 to Danny Jonigk.\nMax-Planck Schools Matter to Life to Marius Reichardt, Tim Salditt.\nhttp://dx.doi.org/10.13039/501100001659Deutsche Forschungsgemeinschaft EXC 2067/1-390729940 to Tim Salditt.\nBotnar Research Center for Child Health BRCCH to Alexandar Tzankov.\n\nAcknowledgements\nWe thank Maximilian Ackermann and Florian Länger for their helpful suggestions, Patrick Zardo for providing control specimen, Emily Brouwer for help in sample preparation, Bastian Hartmann and Jan Goemann for technical help with instrumentation and IT, and Jakob Koch for help in segmentation. It is also our pleasure to acknowledge DESY photon science management for the Covid-19 beamtime call and beamtime.\n\nAdditional information\nCompeting interests\nNo competing interests declared.\nAuthor contributions\nData curation, Software, Formal analysis, Investigation, Visualization, Writing - original draft, Writing - review and editing.\nData curation, Software, Formal analysis, Investigation, Visualization, Writing - original draft, Writing - review and editing.\nData curation, Software, Formal analysis, Investigation, Visualization, Writing - original draft, Writing - review and editing.\nSoftware, Investigation, Methodology, Writing - review and editing.\nResources, Methodology.\nResources, Methodology.\nResources, Validation.\nValidation, Investigation, Visualisation.\nResources, Validation, Writing - original draft.\nConceptualization, Resources, Supervision, Funding acquisition, Validation, Writing - original draft.\nConceptualization, Resources, Data curation, Formal analysis, Supervision, Funding acquisition, Investigation, Methodology, Writing - original draft, Project administration, Writing - review and editing.\nEthics\nHuman subjects: The study was approved by and conducted according to requirements of the ethics committees at the Hannover Medical School (vote Nr. 9022 BO K 2020).\n\nAdditional files\nTransparent reporting form\n\nData availability\nAll datasets were uploaded to zenodo: https://doi.org/10.5281/zenodo.3892637.\nThe following dataset was generated:\nSalditt T Frohn J Eckermann M Reichardt M Osterhoff M Westermeier F Sprung M Tzankov A Kühnel M Jonigk D 2020 3d Virtual Patho-Histology of Lung Tissue from Covid19 Patients based on Phase Contrast X-ray Tomography Zenodo 10.5281/zenodo.3892637\n\nAppendix 1\nMedical information and correlative histology\nBefore recording tomographic scans, tissue sections of 2.5⁢μ⁢m thickness were cut from the top, stained by HE (hematoxylin and eosin) and imaged with a microscope. Appendix 1—figure 1 shows the histological slice of each sample. The imaged section is just above the upper plane of the 3D PC-CT reconstruction volume, but not part of it.\nAppendix 1—figure 1. Microscopic images of HE-stained histological sections of all samples (I–VI).\nHistological slices show comparable morphologies to the virtual slices in Figure 3, which represent different z-position. Scale bars: 400⁢μ⁢m. An overview of different morphological features identified by conventional HE histology and virtual 3D histology is presented in Appendix 1—figure 2, while Appendix 1—figure 3 presents a direct comparison for the same slice. For this purpose, the sample was sectioned and stained after the PC-CT scan. Artery lumen, artery wall, erythrocytes, thrombus, alveolar septum, marcophage, hyaline membrane and black granules (anthracosis) are shown in Appendix 1—figure 2 for both imaging methods. Contrast of the hyaline membrane is homogenous in both modalities, facilitating idetification and segmentation. Erythrocytes are easily recognized by eye in the conventional histology image, due to the HE staining, but less well distinguished by virtual histology. This results in a difficult differentiation between thrombus and blood stasis as well as a difficult identification of blood capillaries in the alveolar septum. Importantly, however, feature identification can be confirmed by correlative 2d and 3D histology on the same section, as exemplified by Appendix 1—figure 3.\nAppendix 1—figure 2. Comparison of morphological features between conventional HE histology and virtual histology.\nArtery lumen, artery wall, erythrocytes, thrombus, alveolar septum, macrophage, hyaline membrane and anthracotic pigments (i.e. the black granules) are presented on exemplary slices of different samples (I–VI) for conventional (left) and virtual histology (right). Scale bars: left 200⁢μ⁢m, right 100⁢μ⁢m.\nAppendix 1—figure 3. Direct comparison of virtual and HE histology for an identical slice/section.\n(top left) Region of interest of the parallel beam tomogram (Sample II). (bottom left) Corresponding HE stained histology slice. Thrombi and erythrocytes can be identified in both imaging modalities. The red square marks the position of the zoom tomogram, for which the corresponding slice is shown on the right. Scale bars: 50⁢μ⁢m.\n\nAppendix 2\nResolution\nResolution estimates are challenging for tissue reconstructions due to the absence of sharp edges or features of well defined size. Here, we follow the approach known as Fourier-shell correlation (FSC). Accordingly, an upper bound for the resolution is be obtained in the following way: the CT scan is split into two, and from each half a 3D volume is reconstructed. After registry, that is, mutual alignment of the volumes, the correlation between the two independent reconstructions in Fourier space is plotted as function of spatial frequency. This correlation must not necessarily reflect the pure system resolution, but instead the range of spatial frequencies over which the results are reliable. As such it is not only affected by the system resolution but also by the sample contrast and the noise of the specific scan. Further, it represents an average over all structures, not taking into account that features with stronger/weaker contrast can correspondingly show higher/lower resolution. Here the Fourier operation for FSC was implemented with a Kaiser-Bessel window of 7 pixels. For the parallel-beam data, a central volume of 650 × 650 × 650 voxels was correlated, for the cone-beam data a volume of 685 × 680 × 250 voxels. These sub-volumes were selected to obtain the average values for the tissue while minimizing contributions from paraffin-filled holes. The correlation curves are shown in Appendix 2—figure 1. The intersection of the curve with the half-bit threshold yields the resolution estimate, indicated with dashed black line. Correspondingly, a half-period resolution of 0.71⁢μ⁢m and 0.39⁢μ⁢m (or better) is obtained for the parallel and cone beam dataset, respectively. However, since the splitted dataset resulted in only 721 angles for the reconstruction, i.e. the resolution estimate is severely affected by under-sampling artifacts, and hence can only serve as an upper bound. Appendix 2—figure 1 illustrates the analysis.\nAppendix 2—figure 1. Quantification of the 3D-resolution for the example of sample V.\nFSC analysis was carried out for two independent reconstructions, each from half the projections of the scan.\n\nAppendix 3\nFurther datasets\nControls: The imaging workflow was also applied to hydrated and/or healthy lung tissue as a control. First, overview scans covering the entire samples were recorded. Then, hydrated 1 mm biopsy punches (two for CTRLI, one for CTRLII and CTRLIII, where CTRLII and CTRLIII are from the same patient) were recorded in (1 - parallel beam) configuration. Biopsy punches from CTRLII and CTRLIII were also examined in (2 - cone beam) mode. Appendix 3—figure 1 presents (a) the rendering of the hydrated control, and (b,c) virtual slices through the reconstruction volume, showing the lung parenchymal architecture in healthy control tissue. UA-stained samples: For all Covid-19 patients, autopsies were also treated by UA-staining in order to increase the contrast. Stitched overview scans in (1 - parallel beam) configuration were recorded, similar to Figure 3. From the UA-labeled tissue blocks of patients I, III, IV and V, 1 mm, biopsy punches were than scanned in the same configuration (Figure 4). Using the (2 - cone beam) setup, these samples from I, III and IV were imaged at 8.0⁢keV x-rays, while V was examined at 13.8⁢keV, as shown for Figure 5. Variation of propagation distance: Scans of the unstained tissue block from patient II were recorded at different propagation distances (z12=50, 100 and 125keV) and different x-ray energies (13.3, 13.8, 14.3 and 14.8 keV). In cone-beam configuration, the unstained biopsy punch from patient I was scanned at 13.8⁢keV x-rays. Compact μCT scans: Prior to the synchrotron experiment, some of the samples have been examined with a laboratory phase-contrast μCT-setup in large mm2-sized FOV-configuration (Liquid metal jet source, Kα=9.5keV, pxeff=5μm, z12=1.7m, 1200 projections of 1 s exposure time with a flat panel CMOS detector with 150⁢μ⁢m Gadox-scintillator, PerkinElmer, USA) (Bartels et al., 2013). Metal-staining (here UA) of the lung tissue helped to achieve sufficient contrast, similar to previous μCT-studies of other biological tissues (Müller et al., 2017; Busse et al., 2018; De Clercq et al., 2019). The resulting overview scans could also be correlated well with histological sections.\nAppendix 3—figure 1. Illustrations of control lung tissue (hydrated).\n(a) Volume rendering of the tissue block (0.97×1.00×0.73 mm3) and (b) slice through the volume, examined in PB-configuration. (c) Slice from the cone-beam scan, arrows indicating the structure of a healthy septum. Particularly, macrophages and erythrocytes emerge. Scale bars: (b) 100⁢μ⁢m and (c) 300⁢μ⁢m.\nAppendix 3—figure 2. Screening with a laboratory phase-contrast μCT-setup.\n(UA-stained tissue block, patient I). (a) Histological and (b) correlative virtual slice from laboratory phase-contrast tomography. (c) Volume rendering of the entire tissue block from a similar perspective. Scale bar: 1 mm."}

LitCovid-PD-CLO

LitCovid-PD-CHEBI

{"project":"LitCovid-PD-CHEBI","denotations":[{"id":"T1","span":{"begin":8247,"end":8249},"obj":"Chemical"},{"id":"T2","span":{"begin":8397,"end":8399},"obj":"Chemical"},{"id":"T3","span":{"begin":8572,"end":8574},"obj":"Chemical"},{"id":"T4","span":{"begin":9667,"end":9669},"obj":"Chemical"}],"attributes":[{"id":"A1","pred":"chebi_id","subj":"T1","obj":"http://purl.obolibrary.org/obo/CHEBI_74327"},{"id":"A2","pred":"chebi_id","subj":"T2","obj":"http://purl.obolibrary.org/obo/CHEBI_74327"},{"id":"A3","pred":"chebi_id","subj":"T3","obj":"http://purl.obolibrary.org/obo/CHEBI_74067"},{"id":"A4","pred":"chebi_id","subj":"T4","obj":"http://purl.obolibrary.org/obo/CHEBI_53319"},{"id":"A5","pred":"chebi_id","subj":"T4","obj":"http://purl.obolibrary.org/obo/CHEBI_60686"}],"text":"Funding Information\nThis paper was supported by the following grants:\nhttp://dx.doi.org/10.13039/501100002347Bundesministerium für Bildung und Forschung 05K19MG2 to Tim Salditt.\nhttp://dx.doi.org/10.13039/100010663H2020 European Research Council 771883 to Danny Jonigk.\nMax-Planck Schools Matter to Life to Marius Reichardt, Tim Salditt.\nhttp://dx.doi.org/10.13039/501100001659Deutsche Forschungsgemeinschaft EXC 2067/1-390729940 to Tim Salditt.\nBotnar Research Center for Child Health BRCCH to Alexandar Tzankov.\n\nAcknowledgements\nWe thank Maximilian Ackermann and Florian Länger for their helpful suggestions, Patrick Zardo for providing control specimen, Emily Brouwer for help in sample preparation, Bastian Hartmann and Jan Goemann for technical help with instrumentation and IT, and Jakob Koch for help in segmentation. It is also our pleasure to acknowledge DESY photon science management for the Covid-19 beamtime call and beamtime.\n\nAdditional information\nCompeting interests\nNo competing interests declared.\nAuthor contributions\nData curation, Software, Formal analysis, Investigation, Visualization, Writing - original draft, Writing - review and editing.\nData curation, Software, Formal analysis, Investigation, Visualization, Writing - original draft, Writing - review and editing.\nData curation, Software, Formal analysis, Investigation, Visualization, Writing - original draft, Writing - review and editing.\nSoftware, Investigation, Methodology, Writing - review and editing.\nResources, Methodology.\nResources, Methodology.\nResources, Validation.\nValidation, Investigation, Visualisation.\nResources, Validation, Writing - original draft.\nConceptualization, Resources, Supervision, Funding acquisition, Validation, Writing - original draft.\nConceptualization, Resources, Data curation, Formal analysis, Supervision, Funding acquisition, Investigation, Methodology, Writing - original draft, Project administration, Writing - review and editing.\nEthics\nHuman subjects: The study was approved by and conducted according to requirements of the ethics committees at the Hannover Medical School (vote Nr. 9022 BO K 2020).\n\nAdditional files\nTransparent reporting form\n\nData availability\nAll datasets were uploaded to zenodo: https://doi.org/10.5281/zenodo.3892637.\nThe following dataset was generated:\nSalditt T Frohn J Eckermann M Reichardt M Osterhoff M Westermeier F Sprung M Tzankov A Kühnel M Jonigk D 2020 3d Virtual Patho-Histology of Lung Tissue from Covid19 Patients based on Phase Contrast X-ray Tomography Zenodo 10.5281/zenodo.3892637\n\nAppendix 1\nMedical information and correlative histology\nBefore recording tomographic scans, tissue sections of 2.5⁢μ⁢m thickness were cut from the top, stained by HE (hematoxylin and eosin) and imaged with a microscope. Appendix 1—figure 1 shows the histological slice of each sample. The imaged section is just above the upper plane of the 3D PC-CT reconstruction volume, but not part of it.\nAppendix 1—figure 1. Microscopic images of HE-stained histological sections of all samples (I–VI).\nHistological slices show comparable morphologies to the virtual slices in Figure 3, which represent different z-position. Scale bars: 400⁢μ⁢m. An overview of different morphological features identified by conventional HE histology and virtual 3D histology is presented in Appendix 1—figure 2, while Appendix 1—figure 3 presents a direct comparison for the same slice. For this purpose, the sample was sectioned and stained after the PC-CT scan. Artery lumen, artery wall, erythrocytes, thrombus, alveolar septum, marcophage, hyaline membrane and black granules (anthracosis) are shown in Appendix 1—figure 2 for both imaging methods. Contrast of the hyaline membrane is homogenous in both modalities, facilitating idetification and segmentation. Erythrocytes are easily recognized by eye in the conventional histology image, due to the HE staining, but less well distinguished by virtual histology. This results in a difficult differentiation between thrombus and blood stasis as well as a difficult identification of blood capillaries in the alveolar septum. Importantly, however, feature identification can be confirmed by correlative 2d and 3D histology on the same section, as exemplified by Appendix 1—figure 3.\nAppendix 1—figure 2. Comparison of morphological features between conventional HE histology and virtual histology.\nArtery lumen, artery wall, erythrocytes, thrombus, alveolar septum, macrophage, hyaline membrane and anthracotic pigments (i.e. the black granules) are presented on exemplary slices of different samples (I–VI) for conventional (left) and virtual histology (right). Scale bars: left 200⁢μ⁢m, right 100⁢μ⁢m.\nAppendix 1—figure 3. Direct comparison of virtual and HE histology for an identical slice/section.\n(top left) Region of interest of the parallel beam tomogram (Sample II). (bottom left) Corresponding HE stained histology slice. Thrombi and erythrocytes can be identified in both imaging modalities. The red square marks the position of the zoom tomogram, for which the corresponding slice is shown on the right. Scale bars: 50⁢μ⁢m.\n\nAppendix 2\nResolution\nResolution estimates are challenging for tissue reconstructions due to the absence of sharp edges or features of well defined size. Here, we follow the approach known as Fourier-shell correlation (FSC). Accordingly, an upper bound for the resolution is be obtained in the following way: the CT scan is split into two, and from each half a 3D volume is reconstructed. After registry, that is, mutual alignment of the volumes, the correlation between the two independent reconstructions in Fourier space is plotted as function of spatial frequency. This correlation must not necessarily reflect the pure system resolution, but instead the range of spatial frequencies over which the results are reliable. As such it is not only affected by the system resolution but also by the sample contrast and the noise of the specific scan. Further, it represents an average over all structures, not taking into account that features with stronger/weaker contrast can correspondingly show higher/lower resolution. Here the Fourier operation for FSC was implemented with a Kaiser-Bessel window of 7 pixels. For the parallel-beam data, a central volume of 650 × 650 × 650 voxels was correlated, for the cone-beam data a volume of 685 × 680 × 250 voxels. These sub-volumes were selected to obtain the average values for the tissue while minimizing contributions from paraffin-filled holes. The correlation curves are shown in Appendix 2—figure 1. The intersection of the curve with the half-bit threshold yields the resolution estimate, indicated with dashed black line. Correspondingly, a half-period resolution of 0.71⁢μ⁢m and 0.39⁢μ⁢m (or better) is obtained for the parallel and cone beam dataset, respectively. However, since the splitted dataset resulted in only 721 angles for the reconstruction, i.e. the resolution estimate is severely affected by under-sampling artifacts, and hence can only serve as an upper bound. Appendix 2—figure 1 illustrates the analysis.\nAppendix 2—figure 1. Quantification of the 3D-resolution for the example of sample V.\nFSC analysis was carried out for two independent reconstructions, each from half the projections of the scan.\n\nAppendix 3\nFurther datasets\nControls: The imaging workflow was also applied to hydrated and/or healthy lung tissue as a control. First, overview scans covering the entire samples were recorded. Then, hydrated 1 mm biopsy punches (two for CTRLI, one for CTRLII and CTRLIII, where CTRLII and CTRLIII are from the same patient) were recorded in (1 - parallel beam) configuration. Biopsy punches from CTRLII and CTRLIII were also examined in (2 - cone beam) mode. Appendix 3—figure 1 presents (a) the rendering of the hydrated control, and (b,c) virtual slices through the reconstruction volume, showing the lung parenchymal architecture in healthy control tissue. UA-stained samples: For all Covid-19 patients, autopsies were also treated by UA-staining in order to increase the contrast. Stitched overview scans in (1 - parallel beam) configuration were recorded, similar to Figure 3. From the UA-labeled tissue blocks of patients I, III, IV and V, 1 mm, biopsy punches were than scanned in the same configuration (Figure 4). Using the (2 - cone beam) setup, these samples from I, III and IV were imaged at 8.0⁢keV x-rays, while V was examined at 13.8⁢keV, as shown for Figure 5. Variation of propagation distance: Scans of the unstained tissue block from patient II were recorded at different propagation distances (z12=50, 100 and 125keV) and different x-ray energies (13.3, 13.8, 14.3 and 14.8 keV). In cone-beam configuration, the unstained biopsy punch from patient I was scanned at 13.8⁢keV x-rays. Compact μCT scans: Prior to the synchrotron experiment, some of the samples have been examined with a laboratory phase-contrast μCT-setup in large mm2-sized FOV-configuration (Liquid metal jet source, Kα=9.5keV, pxeff=5μm, z12=1.7m, 1200 projections of 1 s exposure time with a flat panel CMOS detector with 150⁢μ⁢m Gadox-scintillator, PerkinElmer, USA) (Bartels et al., 2013). Metal-staining (here UA) of the lung tissue helped to achieve sufficient contrast, similar to previous μCT-studies of other biological tissues (Müller et al., 2017; Busse et al., 2018; De Clercq et al., 2019). The resulting overview scans could also be correlated well with histological sections.\nAppendix 3—figure 1. Illustrations of control lung tissue (hydrated).\n(a) Volume rendering of the tissue block (0.97×1.00×0.73 mm3) and (b) slice through the volume, examined in PB-configuration. (c) Slice from the cone-beam scan, arrows indicating the structure of a healthy septum. Particularly, macrophages and erythrocytes emerge. Scale bars: (b) 100⁢μ⁢m and (c) 300⁢μ⁢m.\nAppendix 3—figure 2. Screening with a laboratory phase-contrast μCT-setup.\n(UA-stained tissue block, patient I). (a) Histological and (b) correlative virtual slice from laboratory phase-contrast tomography. (c) Volume rendering of the entire tissue block from a similar perspective. Scale bar: 1 mm."}

LitCovid-PD-GO-BP

{"project":"LitCovid-PD-GO-BP","denotations":[{"id":"T52","span":{"begin":812,"end":824},"obj":"http://purl.obolibrary.org/obo/GO_0035282"},{"id":"T53","span":{"begin":3792,"end":3804},"obj":"http://purl.obolibrary.org/obo/GO_0035282"}],"text":"Funding Information\nThis paper was supported by the following grants:\nhttp://dx.doi.org/10.13039/501100002347Bundesministerium für Bildung und Forschung 05K19MG2 to Tim Salditt.\nhttp://dx.doi.org/10.13039/100010663H2020 European Research Council 771883 to Danny Jonigk.\nMax-Planck Schools Matter to Life to Marius Reichardt, Tim Salditt.\nhttp://dx.doi.org/10.13039/501100001659Deutsche Forschungsgemeinschaft EXC 2067/1-390729940 to Tim Salditt.\nBotnar Research Center for Child Health BRCCH to Alexandar Tzankov.\n\nAcknowledgements\nWe thank Maximilian Ackermann and Florian Länger for their helpful suggestions, Patrick Zardo for providing control specimen, Emily Brouwer for help in sample preparation, Bastian Hartmann and Jan Goemann for technical help with instrumentation and IT, and Jakob Koch for help in segmentation. It is also our pleasure to acknowledge DESY photon science management for the Covid-19 beamtime call and beamtime.\n\nAdditional information\nCompeting interests\nNo competing interests declared.\nAuthor contributions\nData curation, Software, Formal analysis, Investigation, Visualization, Writing - original draft, Writing - review and editing.\nData curation, Software, Formal analysis, Investigation, Visualization, Writing - original draft, Writing - review and editing.\nData curation, Software, Formal analysis, Investigation, Visualization, Writing - original draft, Writing - review and editing.\nSoftware, Investigation, Methodology, Writing - review and editing.\nResources, Methodology.\nResources, Methodology.\nResources, Validation.\nValidation, Investigation, Visualisation.\nResources, Validation, Writing - original draft.\nConceptualization, Resources, Supervision, Funding acquisition, Validation, Writing - original draft.\nConceptualization, Resources, Data curation, Formal analysis, Supervision, Funding acquisition, Investigation, Methodology, Writing - original draft, Project administration, Writing - review and editing.\nEthics\nHuman subjects: The study was approved by and conducted according to requirements of the ethics committees at the Hannover Medical School (vote Nr. 9022 BO K 2020).\n\nAdditional files\nTransparent reporting form\n\nData availability\nAll datasets were uploaded to zenodo: https://doi.org/10.5281/zenodo.3892637.\nThe following dataset was generated:\nSalditt T Frohn J Eckermann M Reichardt M Osterhoff M Westermeier F Sprung M Tzankov A Kühnel M Jonigk D 2020 3d Virtual Patho-Histology of Lung Tissue from Covid19 Patients based on Phase Contrast X-ray Tomography Zenodo 10.5281/zenodo.3892637\n\nAppendix 1\nMedical information and correlative histology\nBefore recording tomographic scans, tissue sections of 2.5⁢μ⁢m thickness were cut from the top, stained by HE (hematoxylin and eosin) and imaged with a microscope. Appendix 1—figure 1 shows the histological slice of each sample. The imaged section is just above the upper plane of the 3D PC-CT reconstruction volume, but not part of it.\nAppendix 1—figure 1. Microscopic images of HE-stained histological sections of all samples (I–VI).\nHistological slices show comparable morphologies to the virtual slices in Figure 3, which represent different z-position. Scale bars: 400⁢μ⁢m. An overview of different morphological features identified by conventional HE histology and virtual 3D histology is presented in Appendix 1—figure 2, while Appendix 1—figure 3 presents a direct comparison for the same slice. For this purpose, the sample was sectioned and stained after the PC-CT scan. Artery lumen, artery wall, erythrocytes, thrombus, alveolar septum, marcophage, hyaline membrane and black granules (anthracosis) are shown in Appendix 1—figure 2 for both imaging methods. Contrast of the hyaline membrane is homogenous in both modalities, facilitating idetification and segmentation. Erythrocytes are easily recognized by eye in the conventional histology image, due to the HE staining, but less well distinguished by virtual histology. This results in a difficult differentiation between thrombus and blood stasis as well as a difficult identification of blood capillaries in the alveolar septum. Importantly, however, feature identification can be confirmed by correlative 2d and 3D histology on the same section, as exemplified by Appendix 1—figure 3.\nAppendix 1—figure 2. Comparison of morphological features between conventional HE histology and virtual histology.\nArtery lumen, artery wall, erythrocytes, thrombus, alveolar septum, macrophage, hyaline membrane and anthracotic pigments (i.e. the black granules) are presented on exemplary slices of different samples (I–VI) for conventional (left) and virtual histology (right). Scale bars: left 200⁢μ⁢m, right 100⁢μ⁢m.\nAppendix 1—figure 3. Direct comparison of virtual and HE histology for an identical slice/section.\n(top left) Region of interest of the parallel beam tomogram (Sample II). (bottom left) Corresponding HE stained histology slice. Thrombi and erythrocytes can be identified in both imaging modalities. The red square marks the position of the zoom tomogram, for which the corresponding slice is shown on the right. Scale bars: 50⁢μ⁢m.\n\nAppendix 2\nResolution\nResolution estimates are challenging for tissue reconstructions due to the absence of sharp edges or features of well defined size. Here, we follow the approach known as Fourier-shell correlation (FSC). Accordingly, an upper bound for the resolution is be obtained in the following way: the CT scan is split into two, and from each half a 3D volume is reconstructed. After registry, that is, mutual alignment of the volumes, the correlation between the two independent reconstructions in Fourier space is plotted as function of spatial frequency. This correlation must not necessarily reflect the pure system resolution, but instead the range of spatial frequencies over which the results are reliable. As such it is not only affected by the system resolution but also by the sample contrast and the noise of the specific scan. Further, it represents an average over all structures, not taking into account that features with stronger/weaker contrast can correspondingly show higher/lower resolution. Here the Fourier operation for FSC was implemented with a Kaiser-Bessel window of 7 pixels. For the parallel-beam data, a central volume of 650 × 650 × 650 voxels was correlated, for the cone-beam data a volume of 685 × 680 × 250 voxels. These sub-volumes were selected to obtain the average values for the tissue while minimizing contributions from paraffin-filled holes. The correlation curves are shown in Appendix 2—figure 1. The intersection of the curve with the half-bit threshold yields the resolution estimate, indicated with dashed black line. Correspondingly, a half-period resolution of 0.71⁢μ⁢m and 0.39⁢μ⁢m (or better) is obtained for the parallel and cone beam dataset, respectively. However, since the splitted dataset resulted in only 721 angles for the reconstruction, i.e. the resolution estimate is severely affected by under-sampling artifacts, and hence can only serve as an upper bound. Appendix 2—figure 1 illustrates the analysis.\nAppendix 2—figure 1. Quantification of the 3D-resolution for the example of sample V.\nFSC analysis was carried out for two independent reconstructions, each from half the projections of the scan.\n\nAppendix 3\nFurther datasets\nControls: The imaging workflow was also applied to hydrated and/or healthy lung tissue as a control. First, overview scans covering the entire samples were recorded. Then, hydrated 1 mm biopsy punches (two for CTRLI, one for CTRLII and CTRLIII, where CTRLII and CTRLIII are from the same patient) were recorded in (1 - parallel beam) configuration. Biopsy punches from CTRLII and CTRLIII were also examined in (2 - cone beam) mode. Appendix 3—figure 1 presents (a) the rendering of the hydrated control, and (b,c) virtual slices through the reconstruction volume, showing the lung parenchymal architecture in healthy control tissue. UA-stained samples: For all Covid-19 patients, autopsies were also treated by UA-staining in order to increase the contrast. Stitched overview scans in (1 - parallel beam) configuration were recorded, similar to Figure 3. From the UA-labeled tissue blocks of patients I, III, IV and V, 1 mm, biopsy punches were than scanned in the same configuration (Figure 4). Using the (2 - cone beam) setup, these samples from I, III and IV were imaged at 8.0⁢keV x-rays, while V was examined at 13.8⁢keV, as shown for Figure 5. Variation of propagation distance: Scans of the unstained tissue block from patient II were recorded at different propagation distances (z12=50, 100 and 125keV) and different x-ray energies (13.3, 13.8, 14.3 and 14.8 keV). In cone-beam configuration, the unstained biopsy punch from patient I was scanned at 13.8⁢keV x-rays. Compact μCT scans: Prior to the synchrotron experiment, some of the samples have been examined with a laboratory phase-contrast μCT-setup in large mm2-sized FOV-configuration (Liquid metal jet source, Kα=9.5keV, pxeff=5μm, z12=1.7m, 1200 projections of 1 s exposure time with a flat panel CMOS detector with 150⁢μ⁢m Gadox-scintillator, PerkinElmer, USA) (Bartels et al., 2013). Metal-staining (here UA) of the lung tissue helped to achieve sufficient contrast, similar to previous μCT-studies of other biological tissues (Müller et al., 2017; Busse et al., 2018; De Clercq et al., 2019). The resulting overview scans could also be correlated well with histological sections.\nAppendix 3—figure 1. Illustrations of control lung tissue (hydrated).\n(a) Volume rendering of the tissue block (0.97×1.00×0.73 mm3) and (b) slice through the volume, examined in PB-configuration. (c) Slice from the cone-beam scan, arrows indicating the structure of a healthy septum. Particularly, macrophages and erythrocytes emerge. Scale bars: (b) 100⁢μ⁢m and (c) 300⁢μ⁢m.\nAppendix 3—figure 2. Screening with a laboratory phase-contrast μCT-setup.\n(UA-stained tissue block, patient I). (a) Histological and (b) correlative virtual slice from laboratory phase-contrast tomography. (c) Volume rendering of the entire tissue block from a similar perspective. Scale bar: 1 mm."}

LitCovid-sentences

LitCovid-PubTator

PMC:7473770 / 43852-54017 JSONTXT

Annnotations TAB JSON ListView MergeView

PMC:7473770 / 43852-54017 JSON TXT