Help Privacy Policy Disclaimer
  Advanced SearchBrowse


  Automatic lung segmentation and quantification of aeration in computed tomography of the chest using 3D transfer learning

Maiello, L., Ball, L., Micali, M., Iannuzzi, F., Scherf, N., Hoffmann, R.-T., et al. (2022). Automatic lung segmentation and quantification of aeration in computed tomography of the chest using 3D transfer learning. Frontiers in Physiology, 12: 725865. doi:10.3389/fphys.2021.725865.

Item is


show Files
hide Files
Maiello_2022.pdf (Publisher version), 3MB
MIME-Type / Checksum:
application/pdf / [MD5]
Technical Metadata:
Copyright Date:
Copyright Info:




Maiello, Lorenzo1, 2, Author
Ball, Lorenzo2, Author
Micali, Marco2, Author
Iannuzzi, Francesca2, Author
Scherf, Nico3, Author           
Hoffmann, Ralf-Thorsten4, Author
Gama de Abreu, Marcelo1, 5, 6, Author
Pelosi, Paolo2, Author
Huhle, Robert1, Author
1Pulmonary Engineering Group, Department of Anaesthesiology and Intensive Care Therapy, University Hospital Carl Gustav Carus, Dresden, Germany, ou_persistent22              
2Department of Surgical Sciences and Integrated Diagnostics, University of Genoa, Italy, ou_persistent22              
3Method and Development Group Neural Data Science and Statistical Computing, MPI for Human Cognitive and Brain Sciences, Max Planck Society, ou_3282987              
4Department of Diagnostic and Interventional Radiology, University Hospital Carl Gustav Carus, Dresden, Germany, ou_persistent22              
5Department of Intensive Care and Resuscitation, Anesthesiology Institute, Cleveland Clinic, OH, USA, ou_persistent22              
6Department of Outcomes Research, Anesthesiology Institute, Cleveland Clinic, OH, USA, ou_persistent22              


Free keywords: uNet; COVID-19; Lung segmentation; ARDS; Jaccard index; Deep learning; Transfer learning; Lung recruitment
 Abstract: Background: Identification of lung parenchyma on computer tomographic (CT) scans in the research setting is done semi-automatically and requires cumbersome manual correction. This is especially true in pathological conditions, hindering the clinical application of aeration compartment (AC) analysis. Deep learning based algorithms have lately been shown to be reliable and time-efficient in segmenting pathologic lungs. In this contribution, we thus propose a novel 3D transfer learning based approach to quantify lung volumes, aeration compartments and lung recruitability.

Methods: Two convolutional neural networks developed for biomedical image segmentation (uNet), with different resolutions and fields of view, were implemented using Matlab. Training and evaluation was done on 180 scans of 18 pigs in experimental ARDS (u2NetPig) and on a clinical data set of 150 scans from 58 ICU patients with lung conditions varying from healthy, to COPD, to ARDS and COVID-19 (u2NetHuman). One manual segmentations (MS) was available for each scan, being a consensus by two experts. Transfer learning was then applied to train u2NetPig on the clinical data set generating u2NetTransfer. General segmentation quality was quantified using the Jaccard index (JI) and the Boundary Function score (BF). The slope between JI or BF and relative volume of non-aerated compartment (SJI and SBF, respectively) was calculated over data sets to assess robustness toward non-aerated lung regions. Additionally, the relative volume of ACs and lung volumes (LV) were compared between automatic and MS.

Results: On the experimental data set, u2NetPig resulted in JI = 0.892 [0.88 : 091] (median [inter-quartile range]), BF = 0.995 [0.98 : 1.0] and slopes SJI = −0.2 {95% conf. int. −0.23 : −0.16} and SBF = −0.1 {−0.5 : −0.06}. u2NetHuman showed similar performance compared to u2NetPig in JI, BF but with reduced robustness SJI = −0.29 {−0.36 : −0.22} and SBF = −0.43 {−0.54 : −0.31}. Transfer learning improved overall JI = 0.92 [0.88 : 0.94], P < 0.001, but reduced robustness SJI = −0.46 {−0.52 : −0.40}, and affected neither BF = 0.96 [0.91 : 0.98] nor SBF = −0.48 {−0.59 : −0.36}. u2NetTransfer improved JI compared to u2NetHuman in segmenting healthy (P = 0.008), ARDS (P < 0.001) and COPD (P = 0.004) patients but not in COVID-19 patients (P = 0.298). ACs and LV determined using u2NetTransfer segmentations exhibited < 5% volume difference compared to MS.

Conclusion: Compared to manual segmentations, automatic uNet based 3D lung segmentation provides acceptable quality for both clinical and scientific purposes in the quantification of lung volumes, aeration compartments, and recruitability.


Language(s): eng - English
 Dates: 2021-06-152021-12-212022-02-04
 Publication Status: Published online
 Pages: -
 Publishing info: -
 Table of Contents: -
 Rev. Type: -
 Identifiers: DOI: 10.3389/fphys.2021.725865
Other: eCollection 2021
PMID: 35185592
PMC: PMC8854801
 Degree: -



Legal Case


Project information

show hide
Project name : -
Grant ID : GA 1256/8-1
Funding program : -
Funding organization : German Research Foundation

Source 1

Title: Frontiers in Physiology
  Other : Front. Physiol.
  Abbreviation : FPHYS
Source Genre: Journal
Publ. Info: Lausanne : Frontiers Research Foundation
Pages: - Volume / Issue: 12 Sequence Number: 725865 Start / End Page: - Identifier: ISSN: 1664-042X
CoNE: https://pure.mpg.de/cone/journals/resource/1664-042X