A Novel Threshold-Based Segmentation Method for Quantification of COVID-19 Lung Abnormalities

Khan, Azrin; Garner, Rachael; Rocca, Marianna La; Salehi, Sana; Duncan, Dominique

doi:10.1007/s11760-022-02183-6

Cited by 17 publications

(16 citation statements)

References 38 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Yet, their thresholding methods recognize advanced quantitative parameters such as histogram shape, measurement space clustering, entropy, or local gray-level surface, which might introduce a benefit compared to the absolute, arbitrary threshold in our study ( 42 ). Similar to our work, Khan et al observed excellent reproduction of AI-based lung involvement segmentation by their threshold-based approaches ( 18 ). After all, the data is yet too scarce to allow for a definitive comparison of our methods, since Khan et al examined only 28 patients ( 18 ).…”

Section: Discussionsupporting

confidence: 89%

“…A further recent study by Khan et al investigated the accuracy of seven established, advanced thresholding methods to segment the affected lung in COVID pneumonia ( 18 ). A parallel to our work is the use of the R231CovidWeb by Hofmanninger et al to obtain the whole-lung segmentation as a starting point ( 31 ).…”

Section: Discussionmentioning

confidence: 99%

“…A parallel to our work is the use of the R231CovidWeb by Hofmanninger et al to obtain the whole-lung segmentation as a starting point ( 31 ). Consecutively, however, Khan et al transformed their data to 8-bit images (256 level greyscale) for further processing, rather than maintaining the established Hounsfield scale ( 18 ). This precludes uncomplicated adoption of their investigated thresholds for clinical use.…”

Section: Discussionmentioning

confidence: 99%

“…All of those findings are apparent by an increase of lung attenuation, when compared to the wellaerated lung tissue. Notably, threshold-based COVID pneumonia segmentation is capable to achieve similar accuracy compared to the recent AI-based approaches (18). Two recent studies investigated arbitrary thresholds for identification of involved lung in COVID-19 (19,20).…”

Section: Introductionmentioning

confidence: 96%

See 3 more Smart Citations

Assessment of COVID-19 lung involvement on computed tomography by deep-learning-, threshold-, and human reader-based approaches—an international, multi-center comparative study

Fervers¹,

Fervers²,

Jaiswal³

et al. 2022

Quant Imaging Med Surg

View full text Add to dashboard Cite

Background: The extent of lung involvement in Coronavirus Disease 2019 (COVID-19) pneumonia, quantified on computed tomography (CT), is an established biomarker for prognosis and guides clinical decision making. The clinical standard is semi-quantitative scoring of lung involvement by an experienced reader. We aim to compare the performance of automated deep-learning-and threshold-based methods to the manual semi-quantitative lung scoring. Further, we aim to investigate an optimal threshold for quantification of involved lung in COVID pneumonia chest CT, using a multi-center dataset. Methods: In total 250 patients were included, 50 consecutive patients with RT-PCR confirmed COVID-19 from our local institutional database, and another 200 patients from four international datasets (n=50 each). Lung involvement was scored semi-quantitatively by three experienced radiologists according to the established chest CT score (CCS) ranging from 0-25. Inter-rater reliability was reported by the intraclass correlation coefficient (ICC). Deep-learning-based segmentation of ground-glass and consolidation was obtained by CT Pulmo Auto Results prototype plugin on IntelliSpace Discovery (Philips Healthcare, The Netherlands). Threshold-based segmentation of involved lung was implemented using an open-source tool for whole-lung segmentation under the presence of severe pathologies (R231CovidWeb, Hofmanninger et al., 2020) and consecutive quantitative assessment of lung attenuation. The best threshold was investigated by training and testing a linear regression of deep-learning and threshold-based results in a five-fold cross validation strategy. Results: Median CCS among 250 evaluated patients was 10 [6-15]. Inter-rater reliability of the CCS was excellent [ICC 0.97 (0.97-0.98)]. Best attenuation threshold for identification of involved lung was −522 HU.While the relationship of deep-learning-and threshold-based quantification was linear and strong (r 2 deep-learning vs. threshold =0.84), both automated quantification methods translated to the semi-quantitative CCS in a nonlinear fashion, with an increasing slope towards higher CCS (r 2 deep-learning vs. CCS = 0.80, r 2 threshold vs. CCS =0.63).

show abstract

Section: Discussionsupporting

confidence: 89%

Section: Discussionmentioning

confidence: 99%

Section: Discussionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 96%

See 2 more Smart Citations

Assessment of COVID-19 lung involvement on computed tomography by deep-learning-, threshold-, and human reader-based approaches—an international, multi-center comparative study

Fervers¹,

Fervers²,

Jaiswal³

et al. 2022

Quant Imaging Med Surg

View full text Add to dashboard Cite

show abstract

“…There were many systematic reviews which showed that ML and DL algorithms can be used for various disease detection and diagnosis processes [21] , [43] , [44] , [45] . These algorithms have proven themselves to be efficient in the detection and classification of various diseases from different types of databases [46] , [47] . Using different DL algorithms, studies have embarked on the identification and differential diagnosis of COVID-19.…”

Section: Discussionmentioning

confidence: 99%

Automated diagnosis of COVID-19 using radiological modalities and Artificial Intelligence functionalities: A retrospective study based on chest HRCT database

Bhattacharjya

Sarma²,

Medhi³

et al. 2023

Biomedical Signal Processing and Control

View full text Add to dashboard Cite

Segmentation of Lung Regions for the Detection of Juxta-Pleura Nodules in CT Scan

Sasidhar

2023

Advances in Intelligent Systems and Computing

View full text Add to dashboard Cite

A Novel Threshold-Based Segmentation Method for Quantification of COVID-19 Lung Abnormalities

Cited by 17 publications

References 38 publications

Assessment of COVID-19 lung involvement on computed tomography by deep-learning-, threshold-, and human reader-based approaches—an international, multi-center comparative study

Assessment of COVID-19 lung involvement on computed tomography by deep-learning-, threshold-, and human reader-based approaches—an international, multi-center comparative study

Automated diagnosis of COVID-19 using radiological modalities and Artificial Intelligence functionalities: A retrospective study based on chest HRCT database

Segmentation of Lung Regions for the Detection of Juxta-Pleura Nodules in CT Scan

Contact Info

Product

Resources

About