A fully automated deep learning pipeline for micro-CT-imaging-based densitometry of lung fibrosis murine models
Abstract Idiopathic pulmonary fibrosis, the archetype of pulmonary fibrosis (PF), is a chronic lung disease of a poor prognosis, characterized by progressively worsening of lung function. Although histology is still the gold standard for PF assessment in preclinical practice, histological data typic...
| Main Authors: | , , , , , , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
BMC
2022-11-01
|
| Series: | Respiratory Research |
| Subjects: | |
| Online Access: | https://doi.org/10.1186/s12931-022-02236-x |
| _version_ | 1811230392810733568 |
|---|---|
| author | Elena Vincenzi Alice Fantazzini Curzio Basso Annalisa Barla Francesca Odone Ludovica Leo Laura Mecozzi Martina Mambrini Erica Ferrini Nicola Sverzellati Franco Fabio Stellari |
| author_facet | Elena Vincenzi Alice Fantazzini Curzio Basso Annalisa Barla Francesca Odone Ludovica Leo Laura Mecozzi Martina Mambrini Erica Ferrini Nicola Sverzellati Franco Fabio Stellari |
| author_sort | Elena Vincenzi |
| collection | DOAJ |
| description | Abstract Idiopathic pulmonary fibrosis, the archetype of pulmonary fibrosis (PF), is a chronic lung disease of a poor prognosis, characterized by progressively worsening of lung function. Although histology is still the gold standard for PF assessment in preclinical practice, histological data typically involve less than 1% of total lung volume and are not amenable to longitudinal studies. A miniaturized version of computed tomography (µCT) has been introduced to radiologically examine lung in preclinical murine models of PF. The linear relationship between X-ray attenuation and tissue density allows lung densitometry on total lung volume. However, the huge density changes caused by PF usually require manual segmentation by trained operators, limiting µCT deployment in preclinical routine. Deep learning approaches have achieved state-of-the-art performance in medical image segmentation. In this work, we propose a fully automated deep learning approach to segment right and left lung on µCT imaging and subsequently derive lung densitometry. Our pipeline first employs a convolutional network (CNN) for pre-processing at low-resolution and then a 2.5D CNN for higher-resolution segmentation, combining computational advantage of 2D and ability to address 3D spatial coherence without compromising accuracy. Finally, lungs are divided into compartments based on air content assessed by density. We validated this pipeline on 72 mice with different grades of PF, achieving a Dice score of 0.967 on test set. Our tests demonstrate that this automated tool allows for rapid and comprehensive analysis of µCT scans of PF murine models, thus laying the ground for its wider exploitation in preclinical settings. |
| first_indexed | 2024-04-12T10:28:41Z |
| format | Article |
| id | doaj.art-da3b3e43dcf74a52b2ccafccd058f3fb |
| institution | Directory Open Access Journal |
| issn | 1465-993X |
| language | English |
| last_indexed | 2024-04-12T10:28:41Z |
| publishDate | 2022-11-01 |
| publisher | BMC |
| record_format | Article |
| series | Respiratory Research |
| spelling | doaj.art-da3b3e43dcf74a52b2ccafccd058f3fb2022-12-22T03:36:55ZengBMCRespiratory Research1465-993X2022-11-0123111410.1186/s12931-022-02236-xA fully automated deep learning pipeline for micro-CT-imaging-based densitometry of lung fibrosis murine modelsElena Vincenzi0Alice Fantazzini1Curzio Basso2Annalisa Barla3Francesca Odone4Ludovica Leo5Laura Mecozzi6Martina Mambrini7Erica Ferrini8Nicola Sverzellati9Franco Fabio Stellari10Department of Computer Science, Bioengineering, Robotics and Systems Engineering, University of GenoaCamelot Biomedical System S.R.LCamelot Biomedical System S.R.LDepartment of Computer Science, Bioengineering, Robotics and Systems Engineering, University of GenoaDepartment of Computer Science, Bioengineering, Robotics and Systems Engineering, University of GenoaDepartment of Medicine and Surgery, University of ParmaDepartment of Medicine and Surgery, University of ParmaDepartment of Veterinary Science, University of ParmaDepartment of Veterinary Science, University of ParmaDepartment of Medicine and Surgery, University of ParmaChiesi Farmaceutici S.P.A, Corporate Pre-Clinical Research and DevelopmentAbstract Idiopathic pulmonary fibrosis, the archetype of pulmonary fibrosis (PF), is a chronic lung disease of a poor prognosis, characterized by progressively worsening of lung function. Although histology is still the gold standard for PF assessment in preclinical practice, histological data typically involve less than 1% of total lung volume and are not amenable to longitudinal studies. A miniaturized version of computed tomography (µCT) has been introduced to radiologically examine lung in preclinical murine models of PF. The linear relationship between X-ray attenuation and tissue density allows lung densitometry on total lung volume. However, the huge density changes caused by PF usually require manual segmentation by trained operators, limiting µCT deployment in preclinical routine. Deep learning approaches have achieved state-of-the-art performance in medical image segmentation. In this work, we propose a fully automated deep learning approach to segment right and left lung on µCT imaging and subsequently derive lung densitometry. Our pipeline first employs a convolutional network (CNN) for pre-processing at low-resolution and then a 2.5D CNN for higher-resolution segmentation, combining computational advantage of 2D and ability to address 3D spatial coherence without compromising accuracy. Finally, lungs are divided into compartments based on air content assessed by density. We validated this pipeline on 72 mice with different grades of PF, achieving a Dice score of 0.967 on test set. Our tests demonstrate that this automated tool allows for rapid and comprehensive analysis of µCT scans of PF murine models, thus laying the ground for its wider exploitation in preclinical settings.https://doi.org/10.1186/s12931-022-02236-xDrug discoveryMicro-computed tomographyDeep learningSegmentationPreclinical |
| spellingShingle | Elena Vincenzi Alice Fantazzini Curzio Basso Annalisa Barla Francesca Odone Ludovica Leo Laura Mecozzi Martina Mambrini Erica Ferrini Nicola Sverzellati Franco Fabio Stellari A fully automated deep learning pipeline for micro-CT-imaging-based densitometry of lung fibrosis murine models Respiratory Research Drug discovery Micro-computed tomography Deep learning Segmentation Preclinical |
| title | A fully automated deep learning pipeline for micro-CT-imaging-based densitometry of lung fibrosis murine models |
| title_full | A fully automated deep learning pipeline for micro-CT-imaging-based densitometry of lung fibrosis murine models |
| title_fullStr | A fully automated deep learning pipeline for micro-CT-imaging-based densitometry of lung fibrosis murine models |
| title_full_unstemmed | A fully automated deep learning pipeline for micro-CT-imaging-based densitometry of lung fibrosis murine models |
| title_short | A fully automated deep learning pipeline for micro-CT-imaging-based densitometry of lung fibrosis murine models |
| title_sort | fully automated deep learning pipeline for micro ct imaging based densitometry of lung fibrosis murine models |
| topic | Drug discovery Micro-computed tomography Deep learning Segmentation Preclinical |
| url | https://doi.org/10.1186/s12931-022-02236-x |
| work_keys_str_mv | AT elenavincenzi afullyautomateddeeplearningpipelineformicroctimagingbaseddensitometryoflungfibrosismurinemodels AT alicefantazzini afullyautomateddeeplearningpipelineformicroctimagingbaseddensitometryoflungfibrosismurinemodels AT curziobasso afullyautomateddeeplearningpipelineformicroctimagingbaseddensitometryoflungfibrosismurinemodels AT annalisabarla afullyautomateddeeplearningpipelineformicroctimagingbaseddensitometryoflungfibrosismurinemodels AT francescaodone afullyautomateddeeplearningpipelineformicroctimagingbaseddensitometryoflungfibrosismurinemodels AT ludovicaleo afullyautomateddeeplearningpipelineformicroctimagingbaseddensitometryoflungfibrosismurinemodels AT lauramecozzi afullyautomateddeeplearningpipelineformicroctimagingbaseddensitometryoflungfibrosismurinemodels AT martinamambrini afullyautomateddeeplearningpipelineformicroctimagingbaseddensitometryoflungfibrosismurinemodels AT ericaferrini afullyautomateddeeplearningpipelineformicroctimagingbaseddensitometryoflungfibrosismurinemodels AT nicolasverzellati afullyautomateddeeplearningpipelineformicroctimagingbaseddensitometryoflungfibrosismurinemodels AT francofabiostellari afullyautomateddeeplearningpipelineformicroctimagingbaseddensitometryoflungfibrosismurinemodels AT elenavincenzi fullyautomateddeeplearningpipelineformicroctimagingbaseddensitometryoflungfibrosismurinemodels AT alicefantazzini fullyautomateddeeplearningpipelineformicroctimagingbaseddensitometryoflungfibrosismurinemodels AT curziobasso fullyautomateddeeplearningpipelineformicroctimagingbaseddensitometryoflungfibrosismurinemodels AT annalisabarla fullyautomateddeeplearningpipelineformicroctimagingbaseddensitometryoflungfibrosismurinemodels AT francescaodone fullyautomateddeeplearningpipelineformicroctimagingbaseddensitometryoflungfibrosismurinemodels AT ludovicaleo fullyautomateddeeplearningpipelineformicroctimagingbaseddensitometryoflungfibrosismurinemodels AT lauramecozzi fullyautomateddeeplearningpipelineformicroctimagingbaseddensitometryoflungfibrosismurinemodels AT martinamambrini fullyautomateddeeplearningpipelineformicroctimagingbaseddensitometryoflungfibrosismurinemodels AT ericaferrini fullyautomateddeeplearningpipelineformicroctimagingbaseddensitometryoflungfibrosismurinemodels AT nicolasverzellati fullyautomateddeeplearningpipelineformicroctimagingbaseddensitometryoflungfibrosismurinemodels AT francofabiostellari fullyautomateddeeplearningpipelineformicroctimagingbaseddensitometryoflungfibrosismurinemodels |