Abstract/Video opens by clicking at the talk title.
Abstract/Video opens by clicking at the talk title.
In-vivo low dose mouse lung imaging using synchrotron phase contrast CT (#393)
Jonas Albers1, 3, Aishwarya Tagat1, Francesca Di Lillo2, Andrea Lorenzon6, Sabina C. Alvarez4, Anna Bergamaschi4, Giuliana Tromba2, Frauke Alves3, 1, 5, Christian Dullin1
1 University Medical Center Goettingen, Institute for Diagnostic and Interventional Radiology, Goettingen, Germany
X-ray imaging is very well suited for lung imaging in mice due to the air functioning as an intrinsic negative contrast agent. Here, we present the first in-vivo x-ray mouse lung imaging study performed at the SYRMEP beamline of the Italian synchrotron “Elettra”. We performed planar imaging to extract lung function parameters, as well as tomographic acquisitions and applied our technique on an ovalbumin induced acute allergic asthma mouse model in comparison to healthy controls.
Female BALB/c mice (4–6 weeks old) were used for an ovalbumin induced asthma mouse model. For x-ray imaging isoflurane anesthetized mice were mounted in upright position (see Figure 1) on a rotating stage. Two different detectors were used for planar as well as tomographic lung imaging. The Photonic Science GSense400 detector (pixel-size 32 µm) and the MOENCH detector (Paul-Scherrer Institute, 25 µm) were utilized at 22 keV in combination with a 1 mm aluminum filter. For planar imaging 1200 projections (Gsense) respectively 4000 projections (MOENCH) were recorded. For tomography 2400 respectively 6000 projections were acquired over a rotation of 540° in 23 seconds. To measure the x-ray dose thermoluminescence dosimeters were implanted into dead mice and imaged in the same way.
By performing free propagation phase contrast lung imaging in living mice using a pixel-size of 32 µm/ 25 µm, we were able to differentiate between asthmatic mice and controls (Fig. 2). The applied x-ray dose of only 80 mGy for a single acquisition was low enough to allow multiple scans in a longitudinal experimental setup. Planar lung function measurements show functional differences between asthmatic and control mice. The severity of the disease model was validated by scoring H&E and PAS stained histological sections. Due to the high flux of the setup frame rates of up to 100 fps were achieved, which allows to extract functional data from the projection images with a higher temporal precision than previously done with a classical in‑vivo microCT. In addition, we will also use this data to improve image quality by reducing motion artifacts.
Inline free propagation phase contrast imaging allows to perform CT in vivo in mouse lungs with a better resolution and lower x-ray dose in comparison to classical absorption-based small animal CT systems. In addition, lung function measurement can be performed at high frame-rates resulting in the extraction of functional alterations of the breathing pattern in an asthmatic lung disease mouse models.
AcknowledgmentThe authors thank all members of the SYRMEP beamline of Elettra and the animal facility of the University of Trieste for making the experiments possible.
Keywords: lung imaging, asthma mouse model, synchrotron imaging, CT imaging
Ultrahigh Resolution Whole-Body Photon-Counting Computed Tomography: A Novel Versatile Tool for Translational Research from Mouse to Man (#25)
Stefan Sawall1, Jan Kuntz1, Carlo Amato1, Laura Klein1, Joscha Maier1, Lukas Thomas Rotkopf2, Eckhard Wehrse2, Danielle Franke4, Nicole Gehrke4, Andreas Briel4, Heinz-Peter Schlemmer2, Christian Herbert Ziener2, Sarah Heinze3, Marc Kachelrieß1
1 German Cancer Research Center (DKFZ), X-Ray Imaging and CT, Heidelberg, Germany
Computed tomography is a valuable tool in clinical practice since it provides cross-sectional images of the specimen under investigation within seconds and a resolution of up to 0.4 mm. Preclinical CT requires dedicated micro-CT systems providing higher spatial resolutions compared to clinical scanners. The need for multiple imaging systems hinders the rapid translation of preclinical findings to clinical practice. This drawback might be overcome by the future introduction of clinical ultrahigh resolution (UHR) photon-counting (PC) CT systems combining preclinical and clinical capabilities.
The prototype of a clinical UHR PCCT (SOMATOM CounT, Siemens) was used for all experiments. The system comprises a conventional energy-integrating (EI) detector and a novel PC detector. While the EI detector provides a pixel size of 0.6 mm in the center of rotation, the PC detector provides a pixel size of 0.25 mm, with future adaptations potentially allowing for 0.125 mm pixel size (fig. 1), and it allows for a quantification of photon energies in up to four distinct energy bins. This acquisition of multi-energy data allows for a multitude of applications, e.g. pseudo-monochromatic imaging. We illustrate the versatile capabilities of UHR PCCT by presenting pilot studies conducted in mice, large animals, human cadavers and patients as well as comparisons with conventional systems.
While conventional clinical CT provides a spatial resolution of up to 13.5 lp/cm (10%-MTF), UHR allows for the acquisition of images with up to 22.4 lp/cm. This corresponds to an object size of 223 µm well suited for the visualization of murine anatomy and finest details of the human body. I.e., all major anatomical structures in mice could be identified (fig. 2) after the administration of a prototype Bismuth-based blood pool agent (nanoPET Pharma, Berlin). Similarly, studies in large animals, human cadavers and patients show a level of detail superior to conventional clinical CT. The intrinsic acquisition of multi-energy data allows for various applications. Most notably, the data allow for the computation of pseudo-monochromatic images. Clinical CT typically offers tube voltages from 70 kV to 150 kV. While preclinical scans could be performed with 70 kV, the computation of pseudo-monochromatic images shows energy levels more appropriate for preclinical settings at lower energies.
Clinical UHR PCCT will boost translational research as it allows for the versatile imaging of specimens from mice to man within a single system while providing clinically well-established applications, e.g. contrast media quantification.
Scan Modes of the Novel Whole-Body Photon-Counting Computed Tomography System
Conventional CT detectors exhibit pixel sizes of about 0.6 mm (left). The novel photon-counting detector provides a pixel size of 0.25 mm (middle). These UHR pixels are actually binned from even smaller pixels with a size of 0.125 mm. Currently, only the 0.25 mm pixels are available due to data transfer rate limitations. However, future updates might allow measurements with a pixel size of down to 0.125 mm.
Translational Research from Mouse to Man
Exemplary pilot experiments conducted using the UHR PCCT system to highlight its versatile capabilities. Top: Acquisition of a mouse (C=200 HU, W=2500 HU) after administration of an experimental Bismuth-based blood pool agent with a Bismuth-overlay in blue. Bottom: Whole-body UHR image of a human cadaver acquired in scope of a forensic study. Sagittal overview (left, C=40 HU, W=300 HU) and the inner ear (right, C=1000 HU, W=3500 HU). Yellow lines in the sagittal reformats indicate the z-position of transversal sections.
Keywords: micro-CT, clinical CT, photon-counting, translational research, contrast agents
Potential of High-Z Elements in Photon-Counting Micro-CT for Optimized Material Decomposition (#255)
Carlo Amato1, 2, Laura Klein1, 2, Joscha Maier1, Stefan Sawall1, Nicole Gehrke3, Danielle Franke3, Spyridon Gkoumas4, Thomas Thüring4, Andreas Briel3, Christian Brönnimann4, Marc Kachelrieß1, 2
1 German Cancer Research Center (DKFZ), Division of X-Ray Imaging and CT (E025), Heidelberg, Germany
The introduction of photon-counting (PC) detectors with adjustable energy thresholds might enable new applications in x-ray computed tomography(CT), e.g. multi-material decomposition. Thanks to their k-edges (Fig.1a), high-Z elements are potential candidates for new contrast media (CM) and might improve material decomposition performance.Herein, different high-Z CM were compared using acquisition protocols optimized for each element in terms of energy thresholds and tube voltage. Improvements in dose normalized contrast-to-noise ratio(CNRD) with respect to the gold standard iodine are measured
Simulations of a mouse-sized (3 cm) phantom with inserts of different CM at the same element concentration (Fig. 1b) were performed for different tube voltages (60-100 kV) relevant for preclinical imaging and for all different combinations of two energy thresholds T0/T1 in 2 keV steps. Spectral detector properties were simulated using a validated spectral response model for a hybrid photon-counting flat detector (Säntis, Dectris Ltd.,Baden, Switzerland). Eight different CM were investigated providing k-edges between 33 keV and 91 keV. An image-based approach was used to perform material decomposition. To evaluate the material decomposition performances, the CNRD was evaluated in the material map of each of the CM. Results were normalized to the CNRD of iodine with a tube voltage of 60 kV.
Energy thresholds maximizing CNRD in material maps were found to be highly dependent on tube voltage and the CM of interest. For each CM, the maximum CNRD is displayed in Fig. 2a as function of the different tube voltages. CNRD maxima were reached by iodine (T=16/32 keV), cerium (T=18/38 keV) and bismuth (T=18/34 keV), at 60 kV. When a higher tube voltage is used, e.g. to decrease beam hardening artifacts or to increase the tube power, bismuth followed by gold outperform the other elements. Independent of the tube voltage, CNRD maxima for bismuth were found at T=18/34 keV. The CNRD improvements of bismuth at 80 kV (Fig. 2b) translate to a radiation dose reduction of 70 % with respect to iodine at the same tube voltage.
Iodine, cerium and bismuth at a tube voltage of 60 kV provided the best performances in material decomposition. At higher tube voltages, heavier elements like Bi and Au can outperform I and Ce and can lead to a significant dose reduction with respect to iodine, given that they can be administered at the desired atomic concentrations. Prototype CM based on Bi are currently being developed for in vivo use (nanoPET Pharma GmbH, Berlin, Germany).
Figure 2 (a) Maximum CNRD as a function of the tube voltage for each investigated contrast medium. (b) Material decomposition of bismuth and water (C200/W2500) with a tube voltage 80 kV and T=18/ 34 keV. At a tube voltage of 80 kV, bone, Au and W appear red as they are a linear combination of soft tissue and Bi.
Keywords: Photon-counting micro-CT, contrast agents, material decomposition
In-vivo contrast enhanced µCT imaging of adult zebrafishes (#20)
Christian Dullin1, Louisa Habich2, Jonas Albers1, Thomas Rittmann3, Giuliana Tromba5, Frauke Alves1, 3, 4, Roland Dosch2
1 University Hospital Goettingen, Inst. for Diagnostic and Interventional Radiology, Goettingen, Germany
On behalf of the x-ray imaging study group of ESMI
Besides mouse and rat, zebrafish is the most commonly applied animal model due to the short generation times, the high fertility and the transparency of its embryos which allows optical 3D imaging. However, the ability to work with adult zebrafish would enable to study diseases that have longer developing times, follow them up over time and/or evaluate treatment efficacy. Imaging in-vivo in adult non-transparent zebrafish is however challenging, mainly due its small size of about 2 cm.
Here we present an in-vivo zebrafish imaging method using contrast enhanced µCT. Adult zebrafish have been anaesthetized with propofol, briefly removed from the tank, injected with 3 µL iodine containing contrast agent (Ultravist 150) into the heart using a microscope, placed back into water and scanned for 4.5 min in water in the QuantumFX in-vivo µCT operated at 90 kV and 200 µA resulting in data sets with a pixel size of 40 µm. In order to validate the imaging results exemplarly ex-vivo scans of zebrafishes stained for 9 days with a 4% iodine in 100% ethanol solution have been performed at the SYRMEP imaging beamline of the Italian synchrotron "Elettra".
While applying a self-made retrospective motion gating and reconstruction software we generated 3D data sets of the fish at different heart states (Figure 1). Using our approach we were able to measure an enlargement of the heart in an Hexabromocyclododecane (HBCD) induced hypertrophy model in-vivo.
In-vivo contrast enhanced microCT imaging in adult zebrafish is feasible and allows for quantification of anatomical modifications of the heart. Further studies are needed to validate if the approximately 2.7 Gy delivered for imaging do not interfere with the well-being of the fish or its fertility. Moreover, the reproducibility of the contrast-agent injection needs to be further increased.
In-vivo microCT imaging results in adult zebrafish of a hypertrophy heart disease model model
Figure 1: shows a virtual cross-section through in-vivo µCT scans of two adult female zebrafishes approx. 5 min after intracardiac injection of 3 µL Ultravist150. From left to right: healthy wild type control (WT), HBCD fish – clearly the heart is enlarged compared to WT. The graphs shows the heart volume normalized by the size of the fish for 5 WT and 3 HBCD fish. As expected the HBCD fishes demonstrate a bigger heart than their WT controls.
Keywords: microCT imaging, contrast agents, zebrafish
New statistical reconstruction method for quantitative imaging in preclinical studies: preliminary results (#148)
Cristobal Martinez Sanchez1, 2, Manuel Desco Menéndez1, 2, 4, Mónica Abella García1, 2, 3
1 Universidad Carlos III de Madrid, Dpto. Bioingeniería e Ingeniería Aeroespacial, Leganes, Spain
The polychromatic nature of the spectra in commercial CT scanners cause an effect known as beam hardening (BH), which hinders the reconstruction of quantitative values. The BH correction in combination with the Hounsfield Units (HU) calibration allow to obtain similar values of only for soft tissue for different kVp but not for bone , introducing a possible confounding factor in bone studies. We propose a new reconstruction method that recovers the tissue density values that are independent of the energy or the acquisition scanner.
We propose a new statistical method that includes the modelling of the polychromatic nature of the spectrum through the BH function, which relates the total attenuation value with the density thickness traversed. This function is obtained in a calibration step with a phantom composed of PMMA and Al-6082, which have equivalent attenuation properties to soft tissue and bone respectively (Figure 1). The algorithm uses this function on the forward model and minimizes the log-likelihood with a Huber function as regularizer.
We can see in top panel of Figure 2 a reduction of the dark bands between the bones and streaks due to the lack of projections with the proposed method. For the conventional reconstruction, the RMSE was 9% of the mean value in bone and 17% in soft tissue. With the proposed reconstruction method these values were reduced to 1.5% in both bone and soft tissue.
We have presented a new statistical reconstruction algorithm that models the polyenergetic nature of the X-ray source with a simple calibration step. Results showed a complete restoration of the density values in simulation data and a complete elimination of the artificats due to beam-hardening effect in real data.
AcknowledgmentThis work has been supported by Ministerio de Ciencia, Innovación y Universidades, Agencia Estatal de Investigación, projects “DPI2016 79075 R AEI/FEDER, UE ”, Instituto de Salud Carlos III, project “DTS17/00122 ”, co funded by European Regional Development Fund (ERDF), “A way of making Europe” Europe”. The CNIC is supported by the Ministerio de Ciencia, Innovación y Universidades and the Pro CNIC Foundation, and is a Severo Ochoa Center of Excellence (SEV 2015 0505).
 Cann, C. E. (1988). Quantitative CT for determination of bone mineral density: a review. Radiology, 166(2), 509-522.
Keywords: X-ray tomography, Quantitave, Polychromatic, Image reconstruction, Beam hardening
Screening of radio-contrast agents for the detection of protein fibrils using X-ray micro-CT (#222)
Michiel Colla1, Caroline Bouillot2, Habiba Tlili3, Thierry Baron3, Brian Metscher4, Fabien Chauveau1
1 Univ. Lyon, Lyon Neuroscience Research Center; CNRS UMR5292; INSERM U1028, Univ. Lyon 1, Lyon, France
Ex vivo X-ray micro-Computed Tomography (micro-CT) of the soft brain tissue using radio-contrast agents (RCA) is emerging as a non-destructive technique to provide high-resolution 3D images, and thus perform “virtual histology” . Various RCA and staining procedures have been tested on brain under normal  or pathological (e.g. glioma, stroke ) conditions. However, no study reported the successful detection of pathological protein aggregates such as amyloid-β or α-synuclein fibrils. This is yet of great interest to study in 3D their prion-like dissemination after brain inoculation .
A first screening step was performed with rats stereotaxically injected with 5µL (200µM) of recombinant α-syn or Aβ pre-formed fibrils. A second step is being performed (in progress) with transgenic APP-PS1 mice .
Staining kinetics and final tissue-to-noise ratio were estimated on low-resolution scans (55µm). Iodine, PTA, gadolinium, eosin and osmium provided the highest level of signal in brain tissue, which reached 3-4 times the one of unstained brain. PTA and osmium penetrated tissue very slowly and complete staining could not always be achieved after one month. In contrast, ioxaglate, iodine and gadolinium yielded homogeneous staining within a week.
Preliminary results suggest that micro-CT detection of endogenous cerebral aggregates of amyloid-β or α-synuclein is feasible. Initial screening in a simplified rat model needs further validation in transgenic mice and identification of the RCA providing the best differentiation over surrounding tissue. The present work intends to provide a new tool for analyzing in 3D the spreading of prion-like deposits in neurodegenerative diseases.
ANR-15-CE18-0026-03; LABEX PRIMES (ANR-11-LABX-0063).
 Albers J, et al. X-ray-Based 3D Virtual Histology-Adding the Next Dimension to Histological Analysis. Mol Imaging Biol. 2018;20(5):732-741.
 Zikmund T, et al. High-contrast differentiation resolution 3D imaging of rodent brain by X-ray computed microtomography. J Inst. 2018;13(02):C02039.
 de Crespigny A, et al. 3D micro-CT imaging of the postmortem brain. J Neurosci Methods. 2008;171(2):207-213 ; Dobrivojević M, et al. Computed microtomography visualization and quantification of mouse ischemic brain lesion by nonionic radio contrast agents. Croat Med J. 2013;54(1):3-11.
 Goedert M. Alzheimer’s and Parkinson’s diseases: The prion concept in relation to assembled Aβ, tau, and α-synuclein. Science. 2015;349(6248):1255555.
 Radde R, et al. Aβ42‐driven cerebral amyloidosis in transgenic mice reveals early and robust pathology. EMBO reports. 2006;7(9):940-946.
Keywords: Contrast-enhanced computer tomography, amyloid, synuclein, radio-contrast agents
X-ray based lung function is a sensitive tool for assessing the progression of weak and strong lung fibrosis in mice (#430)
Amara Khan1, Andrea Markus1, Jonas Albers2, Swen Hülsmann4, Frauke Alves1, 2, 3, Christian Dullin2
1 Max-Planck-Institute for Experimental Medicine, Translational Molecular Imaging, Göttingen, Germany
Idiopathic pulmonary fibrosis (IPF) is a chronic disease of unknown etiology which is characterized by the accumulation of extracellular matrix and fibrillar collagens in the lung. IPF manifests in pulmonary dysfunction and is mainly detected at a late stage. The current diagnosis includes histopathology, biochemical analyses and high resolution CT. Here we applied the novel non-invasive X-ray based lung function (XLF)1,2 measurement approach for sensitive detection of functional parameters of fibrosis in a mouse model of bleomycin induced lung fibrosis.
A mild and strong mouse model of lung fibrosis was produced via intra-tracheal administration of different dosages of bleomycin. In-vivo XLF was recorded one day prior to bleomycin application and on day 7, 14 and 21 after bleomycin administration utilizing low dose planar cinematic x-ray imaging. 2D radiographs were captured for about 30s of the chest movements during 20 breathing cycles, depicting the x-ray transmission of the chest area. Contraction of the lung alters the x-ray transmission over time which can then be parameterized. XLF data were correlated to whole body plethysmography, in vivo and ex-vivo microCT, as well as histology.
XLF was able to detect both, mild and strong fibrosis and monitor it over time. XLF-parameters related to the breathing curve did not show any changes, but significant differences were observed in functional parameters characterizing the dynamics of the lung, such as the anisotropy and the time of inspiration. Other diagnostic tools did not reveal significant changes between mild lung fibrosis and controls: no structural alterations were observed by in vivo microCT, tissue to air content ratio in mild fibrosis was similar to controls in ex vivo CT and whole body plethysmography was unable to identify any differences in the severity of lung fibrosis. Even histology did not show collagen deposition by Masson trichrome staining in mild lung fibrosis. Only XLF was able to show differences at functional level between control and fibrotic mice with mild or strong impact, indicating the superior sensitivity of this imaging approach to all other methods in detecting mild lung fibrosis.
X-ray based lung function is a sensitive tool for diagnosis and monitoring of weak and strong lung fibrosis. Only XLF could show differences at functional level between control and mild fibrotic mice when compared to other detection methods.
AcknowledgmentWe thank Sarah Garbode, Sabine Wolfgramm, Bettina Jeep, Bärbel Heidrich and for excellent technical assistance.
 Dullin, C. Markus, A. Larsson, E. Tromba, G. Hülsmann, S and Alves, F. (2016). X-Ray based Lung Function measurement–a sensitive technique to quantify lung function in allergic airway inﬂammation mouse models. Scientific Reports, 2(6), 36297.
 Markus, A. Borowik, S, Reichardt, M. Tromba, G. Alves, F. and Dullin C. (2017). X-ray-based lung function measurement reveals persistent loss of lung tissue elasticity in mice recovered from allergic airway inflammation. Am J Physiol Lung Cell Mol Physiol 313, L763–L771.
Keywords: XLF, lung fibrosis, microCT
Gas or systemic anesthesia for micro-CT lung imaging? (#576)
Erica Ferrini1, Laura Mecozzi2, Martina Mambrini1, Francesca Ruscitti3, Andrea Grandi3, Sasha Belenkov4, Luisa Corsi3, Gino Villetti3, Fabio F. Stellari3
1 University of Parma, Department of Veterinary Science, Parma, Italy
Gas anesthesia is very popular in preclinical imaging due to its safety profile and versatility. However, a uniform respiratory pattern during the acquisition period is crucial to reduce movement artefacts in lung imaging. Isoflurane (ISO) decreases breathing rate and the protocol is difficult to standardize in different disease models. In this study, isoflurane and the combination of two non-psychotropic systemic agents, Alphaxalone (A) and Dexmedetomidine (D), were used during mCT imaging for the evaluation of antifibrotic drug efficacy in a bleomycin (BLM)-induced lung fibrosis in mice1.
Lung fibrosis was induced in female C57BL/6J mice by double oropharyngeal administration of BLM (days 0 and 4); control (CTR) mice received only saline. Different doses of A and D were intraperitoneally injected to identify the best combination of anesthetics. The selected doses were used for mCT analysis in drug screening studies, in which mice were treated with Drug from day 7 to 21. mCT imaging (Quantum GX, PerkinElmer) was performed at day 21. Lungs were excised for histological analysis to quantify different degrees of fibrosis (mild, moderate and severe lesions)2,3. mCT scans were semi-automatically segmented using Analyze software (Mayo Clinic) and HU clinical ranges were applied4 to define normo- and hypo- aerated regions. The reported data are referred to the end-expiration phase.
The injectable combination of A+D at 30 mg/kg + 0.3 mg/kg was found to be the best option for mCT imaging, since higher doses induced long recovery periods. mCT imaging was performed in CTR, BLM and Drug groups, either using A+D or isoflurane. Normo- and hypo-aerated volumes were quantified and a significant increase (p<0.01) of hypo-aerated tissue was observed in BLM (75%±7 A+D; 65%±12 ISO) compared to CTR group (45%±8 A+D; 34%±7 ISO). On the other hand, the reduction of the hypo-aerated volume after drug treatment was highlighted using both anesthetic protocols and this outcome completely agreed with classical histological analysis (Figure 1). Total lung volumes of A+D treated mice (423±13mm3 CTR; 375±25mm3 BLM) resulted smaller than gas anesthetized mice (501±12mm3 CTR; 462±21mm3 BLM) (p<0.001 within CTR; p<0.05 within BLM), suggesting that systemic anesthesia causes deeper exhalation during the end-expiration phase.
We provided the first evidence that two non-psychotropic agents (Alphaxalone and Dexmedetomidine), which don’t require any approval from national authorities, can be used to perform mCT lung imaging in drug screening experiments as an alternative to isoflurane. Although lung function analysis can be affected, these agents are well tolerated, an important advantage when a high number of animals is recruited.
 Ruscitti, F, Ravanetti, F, Essers, J, et al., 'Longitudinal assessment of bleomycin-induced lung fibrosis by Micro-CT correlates with histological evaluation in mice', Multidiscip Respir Med, 2017;12(1):1-10
 Ashcroft, T, Simpson, JM, Timbrell, V. 'Simple method of estimating severity of pulmonary fibrosis on a numerical scale'. J Clin Pathol. 1988.
 Hübner, RH, Gitter, W, El Mokhtari, NE, et al., 'Standardized quantification of pulmonary fibrosis in histological samples'. Biotechniques. 2008.
 Gattinoni, L, Caironi, P, Pelosi, P, Goodman, LR. 'What has computed tomography taught us about the acute respiratory distress syndrome?', Am J Respir Crit Care Med. 2001;164(9):1701-1711.
Comparison between two independent drug efficacy studies, using systemic or gaseous anesthesia. % hypo-aerated volumes (A. and C.) and Ashcroft distribution of fibrotic lesions (B. and D.) are reported. One-way ANOVA followed by Dunnett’s multiple comparisons test vs BLM group. *p< 0.05; **p< 0.01
Keywords: bleomycin, micro-CT, pulmonary fibrosis, anesthesia