Deep Learning Algorithm Trained with COVID-19 Pneumonia Also Identifies Immune Checkpoint Inhibitor Therapy-Related Pneumonitis
Background: Coronavirus disease 2019 (COVID-19) pneumonia and immune checkpoint inhibitor (ICI) therapy-related pneumonitis share common features. The aim of this study was to determine on chest computed tomography (CT) images whether a deep convolutional neural network algorithm is able to solve th...
| Main Authors: | , , , , , , , , , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2021-02-01
|
| Series: | Cancers |
| Subjects: | |
| Online Access: | https://www.mdpi.com/2072-6694/13/4/652 |
| _version_ | 1797413788666822656 |
|---|---|
| author | Carlo Augusto Mallio Andrea Napolitano Gennaro Castiello Francesco Maria Giordano Pasquale D'Alessio Mario Iozzino Yipeng Sun Silvia Angeletti Marco Russano Daniele Santini Giuseppe Tonini Bruno Beomonte Zobel Bruno Vincenzi Carlo Cosimo Quattrocchi |
| author_facet | Carlo Augusto Mallio Andrea Napolitano Gennaro Castiello Francesco Maria Giordano Pasquale D'Alessio Mario Iozzino Yipeng Sun Silvia Angeletti Marco Russano Daniele Santini Giuseppe Tonini Bruno Beomonte Zobel Bruno Vincenzi Carlo Cosimo Quattrocchi |
| author_sort | Carlo Augusto Mallio |
| collection | DOAJ |
| description | Background: Coronavirus disease 2019 (COVID-19) pneumonia and immune checkpoint inhibitor (ICI) therapy-related pneumonitis share common features. The aim of this study was to determine on chest computed tomography (CT) images whether a deep convolutional neural network algorithm is able to solve the challenge of differential diagnosis between COVID-19 pneumonia and ICI therapy-related pneumonitis. Methods: We enrolled three groups: a pneumonia-free group (<i>n</i> = 30), a COVID-19 group (<i>n</i> = 34), and a group of patients with ICI therapy-related pneumonitis (<i>n</i> = 21). Computed tomography images were analyzed with an artificial intelligence (AI) algorithm based on a deep convolutional neural network structure. Statistical analysis included the Mann–Whitney U test (significance threshold at <i>p</i> < 0.05) and the receiver operating characteristic curve (ROC curve). Results: The algorithm showed low specificity in distinguishing COVID-19 from ICI therapy-related pneumonitis (sensitivity 97.1%, specificity 14.3%, area under the curve (AUC) = 0.62). ICI therapy-related pneumonitis was identified by the AI when compared to pneumonia-free controls (sensitivity = 85.7%, specificity 100%, AUC = 0.97). Conclusions: The deep learning algorithm is not able to distinguish between COVID-19 pneumonia and ICI therapy-related pneumonitis. Awareness must be increased among clinicians about imaging similarities between COVID-19 and ICI therapy-related pneumonitis. ICI therapy-related pneumonitis can be applied as a challenge population for cross-validation to test the robustness of AI models used to analyze interstitial pneumonias of variable etiology. |
| first_indexed | 2024-03-09T05:23:54Z |
| format | Article |
| id | doaj.art-04f0c8c0e6994226b4160329f15c91ed |
| institution | Directory Open Access Journal |
| issn | 2072-6694 |
| language | English |
| last_indexed | 2024-03-09T05:23:54Z |
| publishDate | 2021-02-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Cancers |
| spelling | doaj.art-04f0c8c0e6994226b4160329f15c91ed2023-12-03T12:38:20ZengMDPI AGCancers2072-66942021-02-0113465210.3390/cancers13040652Deep Learning Algorithm Trained with COVID-19 Pneumonia Also Identifies Immune Checkpoint Inhibitor Therapy-Related PneumonitisCarlo Augusto Mallio0Andrea Napolitano1Gennaro Castiello2Francesco Maria Giordano3Pasquale D'Alessio4Mario Iozzino5Yipeng Sun6Silvia Angeletti7Marco Russano8Daniele Santini9Giuseppe Tonini10Bruno Beomonte Zobel11Bruno Vincenzi12Carlo Cosimo Quattrocchi13Departmental Faculty of Medicine and Surgery, Unit of Diagnostic Imaging and Interventional Radiology, Università Campus Bio-Medico di Roma, 00128 Rome, ItalyDepartmental Faculty of Medicine and Surgery, Unit of Medical Oncology, 00128 Rome, ItalyDepartmental Faculty of Medicine and Surgery, Unit of Diagnostic Imaging and Interventional Radiology, Università Campus Bio-Medico di Roma, 00128 Rome, ItalyDepartmental Faculty of Medicine and Surgery, Unit of Diagnostic Imaging and Interventional Radiology, Università Campus Bio-Medico di Roma, 00128 Rome, ItalyDepartmental Faculty of Medicine and Surgery, Unit of Diagnostic Imaging and Interventional Radiology, Università Campus Bio-Medico di Roma, 00128 Rome, ItalyDepartment of Interventional Radiology, S. Maria Goretti Hospital, 04100 Latina, ItalyInfervision Europe GmbH, Mainzer Strasse 75, D-65189 Wiesbaden, GermanyDepartmental Faculty of Medicine and Surgery, Unit of Clinical Laboratory Science, Università Campus Bio-Medico di Roma, 00128 Rome, ItalyDepartmental Faculty of Medicine and Surgery, Unit of Medical Oncology, 00128 Rome, ItalyDepartmental Faculty of Medicine and Surgery, Unit of Medical Oncology, 00128 Rome, ItalyDepartmental Faculty of Medicine and Surgery, Unit of Medical Oncology, 00128 Rome, ItalyDepartmental Faculty of Medicine and Surgery, Unit of Diagnostic Imaging and Interventional Radiology, Università Campus Bio-Medico di Roma, 00128 Rome, ItalyDepartmental Faculty of Medicine and Surgery, Unit of Medical Oncology, 00128 Rome, ItalyDepartmental Faculty of Medicine and Surgery, Unit of Diagnostic Imaging and Interventional Radiology, Università Campus Bio-Medico di Roma, 00128 Rome, ItalyBackground: Coronavirus disease 2019 (COVID-19) pneumonia and immune checkpoint inhibitor (ICI) therapy-related pneumonitis share common features. The aim of this study was to determine on chest computed tomography (CT) images whether a deep convolutional neural network algorithm is able to solve the challenge of differential diagnosis between COVID-19 pneumonia and ICI therapy-related pneumonitis. Methods: We enrolled three groups: a pneumonia-free group (<i>n</i> = 30), a COVID-19 group (<i>n</i> = 34), and a group of patients with ICI therapy-related pneumonitis (<i>n</i> = 21). Computed tomography images were analyzed with an artificial intelligence (AI) algorithm based on a deep convolutional neural network structure. Statistical analysis included the Mann–Whitney U test (significance threshold at <i>p</i> < 0.05) and the receiver operating characteristic curve (ROC curve). Results: The algorithm showed low specificity in distinguishing COVID-19 from ICI therapy-related pneumonitis (sensitivity 97.1%, specificity 14.3%, area under the curve (AUC) = 0.62). ICI therapy-related pneumonitis was identified by the AI when compared to pneumonia-free controls (sensitivity = 85.7%, specificity 100%, AUC = 0.97). Conclusions: The deep learning algorithm is not able to distinguish between COVID-19 pneumonia and ICI therapy-related pneumonitis. Awareness must be increased among clinicians about imaging similarities between COVID-19 and ICI therapy-related pneumonitis. ICI therapy-related pneumonitis can be applied as a challenge population for cross-validation to test the robustness of AI models used to analyze interstitial pneumonias of variable etiology.https://www.mdpi.com/2072-6694/13/4/652artificial intelligenceCOVID-19deep learningchest CTimmune checkpoint inhibitors therapydrug-induced pneumonitis |
| spellingShingle | Carlo Augusto Mallio Andrea Napolitano Gennaro Castiello Francesco Maria Giordano Pasquale D'Alessio Mario Iozzino Yipeng Sun Silvia Angeletti Marco Russano Daniele Santini Giuseppe Tonini Bruno Beomonte Zobel Bruno Vincenzi Carlo Cosimo Quattrocchi Deep Learning Algorithm Trained with COVID-19 Pneumonia Also Identifies Immune Checkpoint Inhibitor Therapy-Related Pneumonitis Cancers artificial intelligence COVID-19 deep learning chest CT immune checkpoint inhibitors therapy drug-induced pneumonitis |
| title | Deep Learning Algorithm Trained with COVID-19 Pneumonia Also Identifies Immune Checkpoint Inhibitor Therapy-Related Pneumonitis |
| title_full | Deep Learning Algorithm Trained with COVID-19 Pneumonia Also Identifies Immune Checkpoint Inhibitor Therapy-Related Pneumonitis |
| title_fullStr | Deep Learning Algorithm Trained with COVID-19 Pneumonia Also Identifies Immune Checkpoint Inhibitor Therapy-Related Pneumonitis |
| title_full_unstemmed | Deep Learning Algorithm Trained with COVID-19 Pneumonia Also Identifies Immune Checkpoint Inhibitor Therapy-Related Pneumonitis |
| title_short | Deep Learning Algorithm Trained with COVID-19 Pneumonia Also Identifies Immune Checkpoint Inhibitor Therapy-Related Pneumonitis |
| title_sort | deep learning algorithm trained with covid 19 pneumonia also identifies immune checkpoint inhibitor therapy related pneumonitis |
| topic | artificial intelligence COVID-19 deep learning chest CT immune checkpoint inhibitors therapy drug-induced pneumonitis |
| url | https://www.mdpi.com/2072-6694/13/4/652 |
| work_keys_str_mv | AT carloaugustomallio deeplearningalgorithmtrainedwithcovid19pneumoniaalsoidentifiesimmunecheckpointinhibitortherapyrelatedpneumonitis AT andreanapolitano deeplearningalgorithmtrainedwithcovid19pneumoniaalsoidentifiesimmunecheckpointinhibitortherapyrelatedpneumonitis AT gennarocastiello deeplearningalgorithmtrainedwithcovid19pneumoniaalsoidentifiesimmunecheckpointinhibitortherapyrelatedpneumonitis AT francescomariagiordano deeplearningalgorithmtrainedwithcovid19pneumoniaalsoidentifiesimmunecheckpointinhibitortherapyrelatedpneumonitis AT pasqualedalessio deeplearningalgorithmtrainedwithcovid19pneumoniaalsoidentifiesimmunecheckpointinhibitortherapyrelatedpneumonitis AT marioiozzino deeplearningalgorithmtrainedwithcovid19pneumoniaalsoidentifiesimmunecheckpointinhibitortherapyrelatedpneumonitis AT yipengsun deeplearningalgorithmtrainedwithcovid19pneumoniaalsoidentifiesimmunecheckpointinhibitortherapyrelatedpneumonitis AT silviaangeletti deeplearningalgorithmtrainedwithcovid19pneumoniaalsoidentifiesimmunecheckpointinhibitortherapyrelatedpneumonitis AT marcorussano deeplearningalgorithmtrainedwithcovid19pneumoniaalsoidentifiesimmunecheckpointinhibitortherapyrelatedpneumonitis AT danielesantini deeplearningalgorithmtrainedwithcovid19pneumoniaalsoidentifiesimmunecheckpointinhibitortherapyrelatedpneumonitis AT giuseppetonini deeplearningalgorithmtrainedwithcovid19pneumoniaalsoidentifiesimmunecheckpointinhibitortherapyrelatedpneumonitis AT brunobeomontezobel deeplearningalgorithmtrainedwithcovid19pneumoniaalsoidentifiesimmunecheckpointinhibitortherapyrelatedpneumonitis AT brunovincenzi deeplearningalgorithmtrainedwithcovid19pneumoniaalsoidentifiesimmunecheckpointinhibitortherapyrelatedpneumonitis AT carlocosimoquattrocchi deeplearningalgorithmtrainedwithcovid19pneumoniaalsoidentifiesimmunecheckpointinhibitortherapyrelatedpneumonitis |