Interpretable Deep Learning for Discriminating Pneumonia from Lung Ultrasounds
Lung ultrasound images have shown great promise to be an operative point-of-care test for the diagnosis of COVID-19 because of the ease of procedure with negligible individual protection equipment, together with relaxed disinfection. Deep learning (DL) is a robust tool for modeling infection pattern...
| Main Authors: | , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2022-11-01
|
| Series: | Mathematics |
| Subjects: | |
| Online Access: | https://www.mdpi.com/2227-7390/10/21/4153 |
| _version_ | 1797467289583353856 |
|---|---|
| author | Mohamed Abdel-Basset Hossam Hawash Khalid Abdulaziz Alnowibet Ali Wagdy Mohamed Karam M. Sallam |
| author_facet | Mohamed Abdel-Basset Hossam Hawash Khalid Abdulaziz Alnowibet Ali Wagdy Mohamed Karam M. Sallam |
| author_sort | Mohamed Abdel-Basset |
| collection | DOAJ |
| description | Lung ultrasound images have shown great promise to be an operative point-of-care test for the diagnosis of COVID-19 because of the ease of procedure with negligible individual protection equipment, together with relaxed disinfection. Deep learning (DL) is a robust tool for modeling infection patterns from medical images; however, the existing COVID-19 detection models are complex and thereby are hard to deploy in frequently used mobile platforms in point-of-care testing. Moreover, most of the COVID-19 detection models in the existing literature on DL are implemented as a black box, hence, they are hard to be interpreted or trusted by the healthcare community. This paper presents a novel interpretable DL framework discriminating COVID-19 infection from other cases of pneumonia and normal cases using ultrasound data of patients. In the proposed framework, novel transformer modules are introduced to model the pathological information from ultrasound frames using an improved window-based multi-head self-attention layer. A convolutional patching module is introduced to transform input frames into latent space rather than partitioning input into patches. A weighted pooling module is presented to score the embeddings of the disease representations obtained from the transformer modules to attend to information that is most valuable for the screening decision. Experimental analysis of the public three-class lung ultrasound dataset (PCUS dataset) demonstrates the discriminative power (Accuracy: 93.4%, F1-score: 93.1%, AUC: 97.5%) of the proposed solution overcoming the competing approaches while maintaining low complexity. The proposed model obtained very promising results in comparison with the rival models. More importantly, it gives explainable outputs therefore, it can serve as a candidate tool for empowering the sustainable diagnosis of COVID-19-like diseases in smart healthcare. |
| first_indexed | 2024-03-09T18:51:30Z |
| format | Article |
| id | doaj.art-05cdeed94c0b49c2bc2776edcb74659e |
| institution | Directory Open Access Journal |
| issn | 2227-7390 |
| language | English |
| last_indexed | 2024-03-09T18:51:30Z |
| publishDate | 2022-11-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Mathematics |
| spelling | doaj.art-05cdeed94c0b49c2bc2776edcb74659e2023-11-24T05:45:43ZengMDPI AGMathematics2227-73902022-11-011021415310.3390/math10214153Interpretable Deep Learning for Discriminating Pneumonia from Lung UltrasoundsMohamed Abdel-Basset0Hossam Hawash1Khalid Abdulaziz Alnowibet2Ali Wagdy Mohamed3Karam M. Sallam4Faculty of Computers and Informatics, Zagazig University, Zagazig 44519, EgyptFaculty of Computers and Informatics, Zagazig University, Zagazig 44519, EgyptStatistics and Operations Research Department, College of Science, King Saud University, P.O. Box 2455, Riyadh 11451, Saudi ArabiaOperations Research Department, Faculty of Graduate Studies for Statistical Research, Cairo University, Giza 12613, EgyptSchool of IT and Systems, University of Canberra, Canberra, ACT 2601, AustraliaLung ultrasound images have shown great promise to be an operative point-of-care test for the diagnosis of COVID-19 because of the ease of procedure with negligible individual protection equipment, together with relaxed disinfection. Deep learning (DL) is a robust tool for modeling infection patterns from medical images; however, the existing COVID-19 detection models are complex and thereby are hard to deploy in frequently used mobile platforms in point-of-care testing. Moreover, most of the COVID-19 detection models in the existing literature on DL are implemented as a black box, hence, they are hard to be interpreted or trusted by the healthcare community. This paper presents a novel interpretable DL framework discriminating COVID-19 infection from other cases of pneumonia and normal cases using ultrasound data of patients. In the proposed framework, novel transformer modules are introduced to model the pathological information from ultrasound frames using an improved window-based multi-head self-attention layer. A convolutional patching module is introduced to transform input frames into latent space rather than partitioning input into patches. A weighted pooling module is presented to score the embeddings of the disease representations obtained from the transformer modules to attend to information that is most valuable for the screening decision. Experimental analysis of the public three-class lung ultrasound dataset (PCUS dataset) demonstrates the discriminative power (Accuracy: 93.4%, F1-score: 93.1%, AUC: 97.5%) of the proposed solution overcoming the competing approaches while maintaining low complexity. The proposed model obtained very promising results in comparison with the rival models. More importantly, it gives explainable outputs therefore, it can serve as a candidate tool for empowering the sustainable diagnosis of COVID-19-like diseases in smart healthcare.https://www.mdpi.com/2227-7390/10/21/4153explainable artificial intelligenceinterpretable deep learningconvolutional networksvision transformersCOVID-19ultrasound image |
| spellingShingle | Mohamed Abdel-Basset Hossam Hawash Khalid Abdulaziz Alnowibet Ali Wagdy Mohamed Karam M. Sallam Interpretable Deep Learning for Discriminating Pneumonia from Lung Ultrasounds Mathematics explainable artificial intelligence interpretable deep learning convolutional networks vision transformers COVID-19 ultrasound image |
| title | Interpretable Deep Learning for Discriminating Pneumonia from Lung Ultrasounds |
| title_full | Interpretable Deep Learning for Discriminating Pneumonia from Lung Ultrasounds |
| title_fullStr | Interpretable Deep Learning for Discriminating Pneumonia from Lung Ultrasounds |
| title_full_unstemmed | Interpretable Deep Learning for Discriminating Pneumonia from Lung Ultrasounds |
| title_short | Interpretable Deep Learning for Discriminating Pneumonia from Lung Ultrasounds |
| title_sort | interpretable deep learning for discriminating pneumonia from lung ultrasounds |
| topic | explainable artificial intelligence interpretable deep learning convolutional networks vision transformers COVID-19 ultrasound image |
| url | https://www.mdpi.com/2227-7390/10/21/4153 |
| work_keys_str_mv | AT mohamedabdelbasset interpretabledeeplearningfordiscriminatingpneumoniafromlungultrasounds AT hossamhawash interpretabledeeplearningfordiscriminatingpneumoniafromlungultrasounds AT khalidabdulazizalnowibet interpretabledeeplearningfordiscriminatingpneumoniafromlungultrasounds AT aliwagdymohamed interpretabledeeplearningfordiscriminatingpneumoniafromlungultrasounds AT karammsallam interpretabledeeplearningfordiscriminatingpneumoniafromlungultrasounds |