Please use this identifier to cite or link to this item:
https://observatorio.fm.usp.br/handle/OPI/40292
Full metadata record
DC Field | Value | Language |
---|---|---|
dc.contributor | Sistema FMUSP-HC: Faculdade de Medicina da Universidade de São Paulo (FMUSP) e Hospital das Clínicas da FMUSP | |
dc.contributor.author | FERREIRA JUNIOR, Jose Raniery | |
dc.contributor.author | CARDENAS, Diego Armando Cardona | |
dc.contributor.author | MORENO, Ramon Alfredo | |
dc.contributor.author | REBELO, Marina de Fatima de Sa | |
dc.contributor.author | KRIEGER, Jose Eduardo | |
dc.contributor.author | GUTIERREZ, Marco Antonio | |
dc.date.accessioned | 2021-06-14T13:23:50Z | - |
dc.date.available | 2021-06-14T13:23:50Z | - |
dc.date.issued | 2020 | |
dc.identifier.citation | 42ND ANNUAL INTERNATIONAL CONFERENCES OF THE IEEE ENGINEERING IN MEDICINE AND BIOLOGY SOCIETY: ENABLING INNOVATIVE TECHNOLOGIES FOR GLOBAL HEALTHCARE EMBC'20, p.1238-1241, 2020 | |
dc.identifier.isbn | 978-1-7281-1990-8 | |
dc.identifier.issn | 1557-170X | |
dc.identifier.uri | https://observatorio.fm.usp.br/handle/OPI/40292 | - |
dc.description.abstract | Pneumonia is one of the leading causes of childhood mortality worldwide. Chest x-ray (CXR) can aid the diagnosis of pneumonia, but in the case of low contrast images, it is important to include computational tools to aid specialists. Deep learning is an alternative because it can identify patterns automatically, even in low-resolution images. We propose herein a convolutional neural network (CNN) architecture with different training strategies towards detecting pneumonia on CXRs and distinguishing its subforms of bacteria and virus. We also evaluated different image pre-processing methods to improve the classification. This study used CXRs from pediatric patients from a public pneumonia CXR dataset. The pre-processing methods evaluated were image cropping and histogram equalization. To classify the images, we adopted the VGG16 CNN and replaced its fully-connected layers with a customized multilayer perceptron. With this architecture, we proposed and evaluated four different training strategies: original CXR image (baseline), chest-cavity-cropped image (A), and histogram-equalized segmented image (B). The last strategy method (C) implemented is based on ensemble between strategies A and B. The performance was assessed by the area under the ROC curve (AUC) with 95% confidence interval (CI), accuracy, sensitivity, specificity, and F1-score. The ensemble model C yielded the highest performances: AUC of 0.97 (CI: 0.96-0.99) to classify pneumonia vs. normal, and AUC of 0.91 (CI: 0.88-0.94) to classify bacterial vs. viral cases. All models that used pre-processed images showed higher AUC than baseline, which used the original CXR image. Image cropping and histogram equalization reduced irrelevant information from the exam, enhanced contrast, and was able to identify fine CXR texture details. The proposed ensemble model increased the representation of inflammatory patterns from bacteria and viruses with few epochs to train the deep CNNs. | eng |
dc.description.sponsorship | Foxconn Brazil and Zerbini Foundation as part of the research project ""Machine Learning in Cardiovascular Medicine"" | |
dc.language.iso | eng | |
dc.publisher | IEEE | eng |
dc.relation.ispartof | 42nd Annual International Conferences of the Ieee Engineering in Medicine and Biology Society: Enabling Innovative Technologies for Global Healthcare Embc'20 | |
dc.relation.ispartofseries | IEEE Engineering in Medicine and Biology Society Conference Proceedings | |
dc.rights | restrictedAccess | eng |
dc.subject.other | diseases | eng |
dc.title | Multi-View Ensemble Convolutional Neural Network to Improve Classification of Pneumonia in Low Contrast Chest X-Ray Images | eng |
dc.type | conferenceObject | eng |
dc.rights.holder | Copyright IEEE | eng |
dc.description.conferencedate | JUL 20-24, 2020 | |
dc.description.conferencelocal | Montreal, CANADA | |
dc.description.conferencename | 42nd Annual International Conference of the IEEE-Engineering-in-Medicine-and-Biology-Society (EMBC) | |
dc.subject.wos | Engineering, Biomedical | eng |
dc.subject.wos | Engineering, Electrical & Electronic | eng |
dc.type.category | proceedings paper | eng |
dc.type.version | publishedVersion | eng |
hcfmusp.description.beginpage | 1238 | |
hcfmusp.description.endpage | 1241 | |
hcfmusp.origem | WOS | |
hcfmusp.origem.id | WOS:000621592201138 | |
hcfmusp.publisher.city | NEW YORK | eng |
hcfmusp.publisher.country | USA | eng |
hcfmusp.relation.reference | Candemir S, 2018, 2018 IEEE LIFE SCIENCES CONFERENCE (LSC), P109, DOI 10.1109/LSC.2018.8572113 | eng |
hcfmusp.relation.reference | Chandra TB, 2020, ADV INTELL SYST, V1022, P21, DOI 10.1007/978-981-32-9088-4_3 | eng |
hcfmusp.relation.reference | Ferreira JR, 2020, INT J COMPUT ASS RAD, V15, P163, DOI 10.1007/s11548-019-02093-y | eng |
hcfmusp.relation.reference | Franquet T, 2018, J THORAC IMAG, V33, P282, DOI 10.1097/RTI.0000000000000347 | eng |
hcfmusp.relation.reference | Irvin J., 2019, ARXIV190107031 | eng |
hcfmusp.relation.reference | Kermany DS, 2018, CELL, V172, P1122, DOI 10.1016/j.cell.2018.02.010 | eng |
hcfmusp.relation.reference | Liang GB, 2020, COMPUT METH PROG BIO, V187, DOI 10.1016/j.cmpb.2019.06.023 | eng |
hcfmusp.relation.reference | Mandell LA, 2007, CLIN INFECT DIS, V44, pS27, DOI 10.1086/511159 | eng |
hcfmusp.relation.reference | Oakden-Rayner L, 2020, ACAD RADIOL, V27, P106, DOI 10.1016/j.acra.2019.10.006 | eng |
hcfmusp.relation.reference | Pazhitnykh I., 2017, LUNG SEGMENTATION 2D | eng |
hcfmusp.relation.reference | Rajpurkar P., 2017, CHEXNET RADIOLOGIST | eng |
hcfmusp.relation.reference | Ronneberger O, 2015, LECT NOTES COMPUT SC, V9351, P234, DOI 10.1007/978-3-319-24574-4_28 | eng |
hcfmusp.relation.reference | Rudan I, 2008, B WORLD HEALTH ORGAN, V86, P408, DOI 10.2471/BLT.07.048769 | eng |
hcfmusp.relation.reference | Russakovsky O, 2015, INT J COMPUT VISION, V115, P211, DOI 10.1007/s11263-015-0816-y | eng |
hcfmusp.relation.reference | Simonyan K., 2014, 14091556 ARXIV, DOI 10.1109/CVPR.2015.7298594 | eng |
hcfmusp.relation.reference | Sousa RT, 2013, PROCEDIA COMPUT SCI, V18, P2579, DOI 10.1016/j.procs.2013.05.444 | eng |
hcfmusp.relation.reference | Wang YB, 2017, PROC CVPR IEEE, P2097, DOI 10.1109/CVPR.2017.226 | eng |
hcfmusp.relation.reference | Zuiderveld K., 1994, GRAPHIC GEMS | eng |
dc.description.index | PubMed | eng |
dc.identifier.eissn | 1558-4615 | |
Appears in Collections: | Comunicações em Eventos - FM/MCP Comunicações em Eventos - HC/InCor Comunicações em Eventos - LIM/13 Comunicações em Eventos - LIM/65 |
Files in This Item:
File | Description | Size | Format | |
---|---|---|---|---|
art_FERREIRA JUNIOR_MultiView_Ensemble_Convolutional_Neural_Network_to_Improve_Classification_2020.PDF Restricted Access | publishedVersion (English) | 1.21 MB | Adobe PDF | View/Open Request a copy |
Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.