Deep Learning-Based Automatic Assessment of Lung Impairment in COVID-19 Pneumonia: Predicting Markers of Hypoxia With Computer Vision
BackgroundHypoxia is a potentially life-threatening condition that can be seen in pneumonia patients.ObjectiveWe aimed to develop and test an automatic assessment of lung impairment in COVID-19 associated pneumonia with machine learning regression models that predict markers of respiratory and cardi...
| Main Authors: | , , , , , , , , , , , , , , , , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Frontiers Media S.A.
2022-07-01
|
| Series: | Frontiers in Medicine |
| Subjects: | |
| Online Access: | https://www.frontiersin.org/articles/10.3389/fmed.2022.882190/full |
| _version_ | 1828369929974317056 |
|---|---|
| author | Yauhen Statsenko Yauhen Statsenko Tetiana Habuza Tetiana Habuza Tatsiana Talako Mikalai Pazniak Elena Likhorad Elena Likhorad Aleh Pazniak Pavel Beliakouski Juri G. Gelovani Juri G. Gelovani Klaus Neidl-Van Gorkom Taleb M. Almansoori Fatmah Al Zahmi Fatmah Al Zahmi Dana Sharif Qandil Nazar Zaki Nazar Zaki Sanaa Elyassami Anna Ponomareva Tom Loney Nerissa Naidoo Guido Hein Huib Mannaerts Guido Hein Huib Mannaerts Jamal Al Koteesh Jamal Al Koteesh Milos R. Ljubisavljevic Karuna M. Das |
| author_facet | Yauhen Statsenko Yauhen Statsenko Tetiana Habuza Tetiana Habuza Tatsiana Talako Mikalai Pazniak Elena Likhorad Elena Likhorad Aleh Pazniak Pavel Beliakouski Juri G. Gelovani Juri G. Gelovani Klaus Neidl-Van Gorkom Taleb M. Almansoori Fatmah Al Zahmi Fatmah Al Zahmi Dana Sharif Qandil Nazar Zaki Nazar Zaki Sanaa Elyassami Anna Ponomareva Tom Loney Nerissa Naidoo Guido Hein Huib Mannaerts Guido Hein Huib Mannaerts Jamal Al Koteesh Jamal Al Koteesh Milos R. Ljubisavljevic Karuna M. Das |
| author_sort | Yauhen Statsenko |
| collection | DOAJ |
| description | BackgroundHypoxia is a potentially life-threatening condition that can be seen in pneumonia patients.ObjectiveWe aimed to develop and test an automatic assessment of lung impairment in COVID-19 associated pneumonia with machine learning regression models that predict markers of respiratory and cardiovascular functioning from radiograms and lung CT.Materials and MethodsWe enrolled a total of 605 COVID-19 cases admitted to Al Ain Hospital from 24 February to 1 July 2020 into the study. The inclusion criteria were as follows: age ≥ 18 years; inpatient admission; PCR positive for SARS-CoV-2; lung CT available at PACS. We designed a CNN-based regression model to predict systemic oxygenation markers from lung CT and 2D diagnostic images of the chest. The 2D images generated by averaging CT scans were analogous to the frontal and lateral view radiograms. The functional (heart and breath rate, blood pressure) and biochemical findings (SpO2, HCO3-, K+, Na+, anion gap, C-reactive protein) served as ground truth.ResultsRadiologic findings in the lungs of COVID-19 patients provide reliable assessments of functional status with clinical utility. If fed to ML models, the sagittal view radiograms reflect dyspnea more accurately than the coronal view radiograms due to the smaller size and the lower model complexity. Mean absolute error of the models trained on single-projection radiograms was approximately 11÷12% and it dropped by 0.5÷1% if both projections were used (11.97 ± 9.23 vs. 11.43 ± 7.51%; p = 0.70). Thus, the ML regression models based on 2D images acquired in multiple planes had slightly better performance. The data blending approach was as efficient as the voting regression technique: 10.90 ± 6.72 vs. 11.96 ± 8.30%, p = 0.94. The models trained on 3D images were more accurate than those on 2D: 8.27 ± 4.13 and 11.75 ± 8.26%, p = 0.14 before lung extraction; 10.66 ± 5.83 and 7.94 ± 4.13%, p = 0.18 after the extraction. The lung extraction boosts 3D model performance unsubstantially (from 8.27 ± 4.13 to 7.94 ± 4.13%; p = 0.82). However, none of the differences between 3D and 2D were statistically significant.ConclusionThe constructed ML algorithms can serve as models of structure-function association and pathophysiologic changes in COVID-19. The algorithms can improve risk evaluation and disease management especially after oxygen therapy that changes functional findings. Thus, the structural assessment of acute lung injury speaks of disease severity. |
| first_indexed | 2024-04-14T06:31:09Z |
| format | Article |
| id | doaj.art-d9a4718dfdc74e2787ad41d6aafdb487 |
| institution | Directory Open Access Journal |
| issn | 2296-858X |
| language | English |
| last_indexed | 2024-04-14T06:31:09Z |
| publishDate | 2022-07-01 |
| publisher | Frontiers Media S.A. |
| record_format | Article |
| series | Frontiers in Medicine |
| spelling | doaj.art-d9a4718dfdc74e2787ad41d6aafdb4872022-12-22T02:07:38ZengFrontiers Media S.A.Frontiers in Medicine2296-858X2022-07-01910.3389/fmed.2022.882190882190Deep Learning-Based Automatic Assessment of Lung Impairment in COVID-19 Pneumonia: Predicting Markers of Hypoxia With Computer VisionYauhen Statsenko0Yauhen Statsenko1Tetiana Habuza2Tetiana Habuza3Tatsiana Talako4Mikalai Pazniak5Elena Likhorad6Elena Likhorad7Aleh Pazniak8Pavel Beliakouski9Juri G. Gelovani10Juri G. Gelovani11Klaus Neidl-Van Gorkom12Taleb M. Almansoori13Fatmah Al Zahmi14Fatmah Al Zahmi15Dana Sharif Qandil16Nazar Zaki17Nazar Zaki18Sanaa Elyassami19Anna Ponomareva20Tom Loney21Nerissa Naidoo22Guido Hein Huib Mannaerts23Guido Hein Huib Mannaerts24Jamal Al Koteesh25Jamal Al Koteesh26Milos R. Ljubisavljevic27Karuna M. Das28Department of Radiology, College of Medicine and Health Sciences, United Arab Emirates University, Al Ain, United Arab EmiratesAbu Dhabi Precision Medicine Virtual Research Institute (AD PM VRI), United Arab Emirates University, Al Ain, United Arab EmiratesDepartment of Computer Science and Software Engineering, College of Information Technology, United Arab Emirates University, Al Ain, United Arab EmiratesBig Data Analytics Center, United Arab Emirates University, Al Ain, United Arab EmiratesDepartment of Radiology, College of Medicine and Health Sciences, United Arab Emirates University, Al Ain, United Arab EmiratesEye Microsurgery Center “Voka”, Minsk, BelarusDepartment of Radiology, College of Medicine and Health Sciences, United Arab Emirates University, Al Ain, United Arab EmiratesEye Microsurgery Center “Voka”, Minsk, BelarusEye Microsurgery Center “Voka”, Minsk, BelarusEye Microsurgery Center “Voka”, Minsk, BelarusBiomedical Engineering Department, College of Engineering, Wayne State University, Detroit, MI, United StatesSiriraj Hospital, Mahidol University, Nakhon Pathom, ThailandDepartment of Radiology, College of Medicine and Health Sciences, United Arab Emirates University, Al Ain, United Arab EmiratesDepartment of Radiology, College of Medicine and Health Sciences, United Arab Emirates University, Al Ain, United Arab EmiratesDepartment of Neurology, Mediclinic Parkview Hospital, Dubai, United Arab EmiratesDepartment of Clinical Science, College of Medicine, Mohammed Bin Rashid University of Medicine and Health Sciences, Dubai, United Arab Emirates0College of Medical Sciences, Ras Al Khaimah Medical Health and Sciences University, Ras Al Khaimah, United Arab EmiratesAbu Dhabi Precision Medicine Virtual Research Institute (AD PM VRI), United Arab Emirates University, Al Ain, United Arab EmiratesDepartment of Computer Science and Software Engineering, College of Information Technology, United Arab Emirates University, Al Ain, United Arab Emirates1Department of Computer Science, Abu Dhabi Polytechnic, Abu Dhabi, United Arab Emirates2Scientific-Research Institute of Medicine and Dentistry, Moscow State University of Medicine and Dentistry, Moscow, Russia3Department of Public Health and Epidemiology, College of Medicine, Mohammed Bin Rashid University of Medicine and Health Sciences, Dubai, United Arab Emirates4Department of Anatomy, College of Medicine, Mohammed Bin Rashid University of Medicine and Health Sciences, Dubai, United Arab Emirates5Department of Surgery, College of Medicine and Health Sciences, United Arab Emirates University, Al Ain, United Arab Emirates6Department of Surgery, Tawam Hospital, Abu Dhabi, United Arab EmiratesDepartment of Radiology, College of Medicine and Health Sciences, United Arab Emirates University, Al Ain, United Arab Emirates7Department of Radiology, Tawam Hospital, Abu Dhabi, United Arab Emirates8Department of Physiology, College of Medicine and Health Sciences, United Arab Emirates University, Al Ain, United Arab EmiratesDepartment of Radiology, College of Medicine and Health Sciences, United Arab Emirates University, Al Ain, United Arab EmiratesBackgroundHypoxia is a potentially life-threatening condition that can be seen in pneumonia patients.ObjectiveWe aimed to develop and test an automatic assessment of lung impairment in COVID-19 associated pneumonia with machine learning regression models that predict markers of respiratory and cardiovascular functioning from radiograms and lung CT.Materials and MethodsWe enrolled a total of 605 COVID-19 cases admitted to Al Ain Hospital from 24 February to 1 July 2020 into the study. The inclusion criteria were as follows: age ≥ 18 years; inpatient admission; PCR positive for SARS-CoV-2; lung CT available at PACS. We designed a CNN-based regression model to predict systemic oxygenation markers from lung CT and 2D diagnostic images of the chest. The 2D images generated by averaging CT scans were analogous to the frontal and lateral view radiograms. The functional (heart and breath rate, blood pressure) and biochemical findings (SpO2, HCO3-, K+, Na+, anion gap, C-reactive protein) served as ground truth.ResultsRadiologic findings in the lungs of COVID-19 patients provide reliable assessments of functional status with clinical utility. If fed to ML models, the sagittal view radiograms reflect dyspnea more accurately than the coronal view radiograms due to the smaller size and the lower model complexity. Mean absolute error of the models trained on single-projection radiograms was approximately 11÷12% and it dropped by 0.5÷1% if both projections were used (11.97 ± 9.23 vs. 11.43 ± 7.51%; p = 0.70). Thus, the ML regression models based on 2D images acquired in multiple planes had slightly better performance. The data blending approach was as efficient as the voting regression technique: 10.90 ± 6.72 vs. 11.96 ± 8.30%, p = 0.94. The models trained on 3D images were more accurate than those on 2D: 8.27 ± 4.13 and 11.75 ± 8.26%, p = 0.14 before lung extraction; 10.66 ± 5.83 and 7.94 ± 4.13%, p = 0.18 after the extraction. The lung extraction boosts 3D model performance unsubstantially (from 8.27 ± 4.13 to 7.94 ± 4.13%; p = 0.82). However, none of the differences between 3D and 2D were statistically significant.ConclusionThe constructed ML algorithms can serve as models of structure-function association and pathophysiologic changes in COVID-19. The algorithms can improve risk evaluation and disease management especially after oxygen therapy that changes functional findings. Thus, the structural assessment of acute lung injury speaks of disease severity.https://www.frontiersin.org/articles/10.3389/fmed.2022.882190/fullblended machine learning modeldeep learningCOVID-19pneumoniaSARC-CoV-2lung structural changes |
| spellingShingle | Yauhen Statsenko Yauhen Statsenko Tetiana Habuza Tetiana Habuza Tatsiana Talako Mikalai Pazniak Elena Likhorad Elena Likhorad Aleh Pazniak Pavel Beliakouski Juri G. Gelovani Juri G. Gelovani Klaus Neidl-Van Gorkom Taleb M. Almansoori Fatmah Al Zahmi Fatmah Al Zahmi Dana Sharif Qandil Nazar Zaki Nazar Zaki Sanaa Elyassami Anna Ponomareva Tom Loney Nerissa Naidoo Guido Hein Huib Mannaerts Guido Hein Huib Mannaerts Jamal Al Koteesh Jamal Al Koteesh Milos R. Ljubisavljevic Karuna M. Das Deep Learning-Based Automatic Assessment of Lung Impairment in COVID-19 Pneumonia: Predicting Markers of Hypoxia With Computer Vision Frontiers in Medicine blended machine learning model deep learning COVID-19 pneumonia SARC-CoV-2 lung structural changes |
| title | Deep Learning-Based Automatic Assessment of Lung Impairment in COVID-19 Pneumonia: Predicting Markers of Hypoxia With Computer Vision |
| title_full | Deep Learning-Based Automatic Assessment of Lung Impairment in COVID-19 Pneumonia: Predicting Markers of Hypoxia With Computer Vision |
| title_fullStr | Deep Learning-Based Automatic Assessment of Lung Impairment in COVID-19 Pneumonia: Predicting Markers of Hypoxia With Computer Vision |
| title_full_unstemmed | Deep Learning-Based Automatic Assessment of Lung Impairment in COVID-19 Pneumonia: Predicting Markers of Hypoxia With Computer Vision |
| title_short | Deep Learning-Based Automatic Assessment of Lung Impairment in COVID-19 Pneumonia: Predicting Markers of Hypoxia With Computer Vision |
| title_sort | deep learning based automatic assessment of lung impairment in covid 19 pneumonia predicting markers of hypoxia with computer vision |
| topic | blended machine learning model deep learning COVID-19 pneumonia SARC-CoV-2 lung structural changes |
| url | https://www.frontiersin.org/articles/10.3389/fmed.2022.882190/full |
| work_keys_str_mv | AT yauhenstatsenko deeplearningbasedautomaticassessmentoflungimpairmentincovid19pneumoniapredictingmarkersofhypoxiawithcomputervision AT yauhenstatsenko deeplearningbasedautomaticassessmentoflungimpairmentincovid19pneumoniapredictingmarkersofhypoxiawithcomputervision AT tetianahabuza deeplearningbasedautomaticassessmentoflungimpairmentincovid19pneumoniapredictingmarkersofhypoxiawithcomputervision AT tetianahabuza deeplearningbasedautomaticassessmentoflungimpairmentincovid19pneumoniapredictingmarkersofhypoxiawithcomputervision AT tatsianatalako deeplearningbasedautomaticassessmentoflungimpairmentincovid19pneumoniapredictingmarkersofhypoxiawithcomputervision AT mikalaipazniak deeplearningbasedautomaticassessmentoflungimpairmentincovid19pneumoniapredictingmarkersofhypoxiawithcomputervision AT elenalikhorad deeplearningbasedautomaticassessmentoflungimpairmentincovid19pneumoniapredictingmarkersofhypoxiawithcomputervision AT elenalikhorad deeplearningbasedautomaticassessmentoflungimpairmentincovid19pneumoniapredictingmarkersofhypoxiawithcomputervision AT alehpazniak deeplearningbasedautomaticassessmentoflungimpairmentincovid19pneumoniapredictingmarkersofhypoxiawithcomputervision AT pavelbeliakouski deeplearningbasedautomaticassessmentoflungimpairmentincovid19pneumoniapredictingmarkersofhypoxiawithcomputervision AT juriggelovani deeplearningbasedautomaticassessmentoflungimpairmentincovid19pneumoniapredictingmarkersofhypoxiawithcomputervision AT juriggelovani deeplearningbasedautomaticassessmentoflungimpairmentincovid19pneumoniapredictingmarkersofhypoxiawithcomputervision AT klausneidlvangorkom deeplearningbasedautomaticassessmentoflungimpairmentincovid19pneumoniapredictingmarkersofhypoxiawithcomputervision AT talebmalmansoori deeplearningbasedautomaticassessmentoflungimpairmentincovid19pneumoniapredictingmarkersofhypoxiawithcomputervision AT fatmahalzahmi deeplearningbasedautomaticassessmentoflungimpairmentincovid19pneumoniapredictingmarkersofhypoxiawithcomputervision AT fatmahalzahmi deeplearningbasedautomaticassessmentoflungimpairmentincovid19pneumoniapredictingmarkersofhypoxiawithcomputervision AT danasharifqandil deeplearningbasedautomaticassessmentoflungimpairmentincovid19pneumoniapredictingmarkersofhypoxiawithcomputervision AT nazarzaki deeplearningbasedautomaticassessmentoflungimpairmentincovid19pneumoniapredictingmarkersofhypoxiawithcomputervision AT nazarzaki deeplearningbasedautomaticassessmentoflungimpairmentincovid19pneumoniapredictingmarkersofhypoxiawithcomputervision AT sanaaelyassami deeplearningbasedautomaticassessmentoflungimpairmentincovid19pneumoniapredictingmarkersofhypoxiawithcomputervision AT annaponomareva deeplearningbasedautomaticassessmentoflungimpairmentincovid19pneumoniapredictingmarkersofhypoxiawithcomputervision AT tomloney deeplearningbasedautomaticassessmentoflungimpairmentincovid19pneumoniapredictingmarkersofhypoxiawithcomputervision AT nerissanaidoo deeplearningbasedautomaticassessmentoflungimpairmentincovid19pneumoniapredictingmarkersofhypoxiawithcomputervision AT guidoheinhuibmannaerts deeplearningbasedautomaticassessmentoflungimpairmentincovid19pneumoniapredictingmarkersofhypoxiawithcomputervision AT guidoheinhuibmannaerts deeplearningbasedautomaticassessmentoflungimpairmentincovid19pneumoniapredictingmarkersofhypoxiawithcomputervision AT jamalalkoteesh deeplearningbasedautomaticassessmentoflungimpairmentincovid19pneumoniapredictingmarkersofhypoxiawithcomputervision AT jamalalkoteesh deeplearningbasedautomaticassessmentoflungimpairmentincovid19pneumoniapredictingmarkersofhypoxiawithcomputervision AT milosrljubisavljevic deeplearningbasedautomaticassessmentoflungimpairmentincovid19pneumoniapredictingmarkersofhypoxiawithcomputervision AT karunamdas deeplearningbasedautomaticassessmentoflungimpairmentincovid19pneumoniapredictingmarkersofhypoxiawithcomputervision |