COVLIAS 1.0: Lung Segmentation in COVID-19 Computed Tomography Scans Using Hybrid Deep Learning Artificial Intelligence Models.
COVID-19
computed tomography
hybrid deep learning
lungs
segmentation
Journal
Diagnostics (Basel, Switzerland)
ISSN: 2075-4418
Titre abrégé: Diagnostics (Basel)
Pays: Switzerland
ID NLM: 101658402
Informations de publication
Date de publication:
04 Aug 2021
04 Aug 2021
Historique:
received:
06
07
2021
revised:
28
07
2021
accepted:
29
07
2021
entrez:
27
8
2021
pubmed:
28
8
2021
medline:
28
8
2021
Statut:
epublish
Résumé
COVID-19 lung segmentation using Computed Tomography (CT) scans is important for the diagnosis of lung severity. The process of automated lung segmentation is challenging due to (a) CT radiation dosage and (b) ground-glass opacities caused by COVID-19. The lung segmentation methodologies proposed in 2020 were semi- or automated but not reliable, accurate, and user-friendly. The proposed study presents a COVID Lung Image Analysis System (COVLIAS 1.0, AtheroPoint™, Roseville, CA, USA) consisting of hybrid deep learning (HDL) models for lung segmentation. The COVLIAS 1.0 consists of three methods based on solo deep learning (SDL) or hybrid deep learning (HDL). SegNet is proposed in the SDL category while VGG-SegNet and ResNet-SegNet are designed under the HDL paradigm. The three proposed AI approaches were benchmarked against the National Institute of Health (NIH)-based conventional segmentation model using fuzzy-connectedness. A cross-validation protocol with a 40:60 ratio between training and testing was designed, with 10% validation data. The ground truth (GT) was manually traced by a radiologist trained personnel. For performance evaluation, nine different criteria were selected to perform the evaluation of SDL or HDL lung segmentation regions and lungs long axis against GT. Using the database of 5000 chest CT images (from 72 patients), COVLIAS 1.0 yielded AUC of The COVLIAS 1.0 system can be applied in real-time for radiology-based clinical settings.
Sections du résumé
BACKGROUND
BACKGROUND
COVID-19 lung segmentation using Computed Tomography (CT) scans is important for the diagnosis of lung severity. The process of automated lung segmentation is challenging due to (a) CT radiation dosage and (b) ground-glass opacities caused by COVID-19. The lung segmentation methodologies proposed in 2020 were semi- or automated but not reliable, accurate, and user-friendly. The proposed study presents a COVID Lung Image Analysis System (COVLIAS 1.0, AtheroPoint™, Roseville, CA, USA) consisting of hybrid deep learning (HDL) models for lung segmentation.
METHODOLOGY
METHODS
The COVLIAS 1.0 consists of three methods based on solo deep learning (SDL) or hybrid deep learning (HDL). SegNet is proposed in the SDL category while VGG-SegNet and ResNet-SegNet are designed under the HDL paradigm. The three proposed AI approaches were benchmarked against the National Institute of Health (NIH)-based conventional segmentation model using fuzzy-connectedness. A cross-validation protocol with a 40:60 ratio between training and testing was designed, with 10% validation data. The ground truth (GT) was manually traced by a radiologist trained personnel. For performance evaluation, nine different criteria were selected to perform the evaluation of SDL or HDL lung segmentation regions and lungs long axis against GT.
RESULTS
RESULTS
Using the database of 5000 chest CT images (from 72 patients), COVLIAS 1.0 yielded AUC of
CONCLUSIONS
CONCLUSIONS
The COVLIAS 1.0 system can be applied in real-time for radiology-based clinical settings.
Identifiants
pubmed: 34441340
pii: diagnostics11081405
doi: 10.3390/diagnostics11081405
pmc: PMC8392426
pii:
doi:
Types de publication
Journal Article
Langues
eng
Références
Annu Int Conf IEEE Eng Med Biol Soc. 2014;2014:1087-90
pubmed: 25570151
Comput Methods Programs Biomed. 2013 Dec;112(3):624-32
pubmed: 23958645
Comput Methods Programs Biomed. 2016 Apr;126:98-109
pubmed: 26830378
Eur Rev Med Pharmacol Sci. 2020 Dec;24(23):12609-12622
pubmed: 33336781
IEEE Trans Ultrason Ferroelectr Freq Control. 2010 May;57(5):1112-24
pubmed: 20442021
Med Phys. 2012 Jul;39(7):4255-64
pubmed: 22830759
Technol Cancer Res Treat. 2014 Aug;13(4):289-301
pubmed: 24206204
Front Med (Lausanne). 2020 Dec 23;7:608525
pubmed: 33425953
Clin Imaging. 2021 Sep;77:244-249
pubmed: 34029929
Comput Methods Programs Biomed. 2018 Mar;155:165-177
pubmed: 29512496
J Med Imaging (Bellingham). 2021 Jan;8(Suppl 1):014502
pubmed: 33912622
Rev Cardiovasc Med. 2020 Dec 30;21(4):541-560
pubmed: 33387999
Comput Biol Med. 2020 Oct;125:103958
pubmed: 32927257
Cardiovasc Diagn Ther. 2020 Aug;10(4):919-938
pubmed: 32968651
J Med Syst. 2018 Apr 10;42(5):92
pubmed: 29637403
BMC Med Imaging. 2021 Feb 9;21(1):19
pubmed: 33557772
Int J Cardiovasc Imaging. 2021 Apr;37(4):1349-1360
pubmed: 33211242
J Stroke. 2018 Sep;20(3):302-320
pubmed: 30309226
Biomed Res Int. 2021 Apr 15;2021:5544742
pubmed: 33954175
Comput Biol Med. 2016 Aug 1;75:217-34
pubmed: 27318571
Cardiovasc Diagn Ther. 2019 Oct;9(5):420-430
pubmed: 31737514
Ultrasonics. 2012 Apr;52(4):508-20
pubmed: 22154208
IEEE Access. 2020 Sep 30;8:179317-179335
pubmed: 34976558
Int J Comput Assist Radiol Surg. 2021 Mar;16(3):423-434
pubmed: 33532975
Cardiovasc Diagn Ther. 2019 Oct;9(5):439-461
pubmed: 31737516
J Med Syst. 2015 Mar;39(3):22
pubmed: 25666926
Comput Methods Programs Biomed. 2012 Dec;108(3):946-60
pubmed: 22658832
Indian Heart J. 2018 Sep - Oct;70(5):649-664
pubmed: 30392503
Ultrasound Med Biol. 2010 Oct;36(10):1616-25
pubmed: 20800947
J Med Syst. 2017 Aug 23;41(10):152
pubmed: 28836045
Comput Biol Med. 2017 Oct 1;89:197-211
pubmed: 28825994
J Med Syst. 2021 Jan 26;45(3):28
pubmed: 33496876
Sheng Wu Yi Xue Gong Cheng Xue Za Zhi. 2021 Apr 25;38(2):379-386
pubmed: 33913299
Echocardiography. 2019 Feb;36(2):345-361
pubmed: 30623485
World J Diabetes. 2021 Mar 15;12(3):215-237
pubmed: 33758644
Mater Today Proc. 2021 Feb 5;:
pubmed: 33564622
Comput Methods Programs Biomed. 2016 May;128:137-58
pubmed: 27040838
Comput Biol Med. 2020 Sep;124:103960
pubmed: 32919186
PeerJ Comput Sci. 2021 Feb 22;7:e368
pubmed: 33817018
J Med Syst. 2012 Jun;36(3):1861-71
pubmed: 21243411
Eur J Radiol. 2019 May;114:14-24
pubmed: 31005165
Front Biosci (Landmark Ed). 2019 Jan 1;24(3):392-426
pubmed: 30468663
Comput Biol Med. 2021 Mar;130:104210
pubmed: 33550068
J Med Syst. 2016 Jun;40(6):142
pubmed: 27114353
J Public Health Res. 2021 Apr 19;10(3):
pubmed: 33876627
IEEE Trans Pattern Anal Mach Intell. 2017 Dec;39(12):2481-2495
pubmed: 28060704
IEEE Trans Neural Netw Learn Syst. 2021 Mar;32(3):932-946
pubmed: 33544680
Acad Radiol. 2020 Dec;27(12):1665-1678
pubmed: 33046370
Comput Biol Med. 2017 Dec 1;91:306-317
pubmed: 29107894