TY - JOUR
T1 - Deep learning representations to support COVID-19 diagnosis on CT-slices
AU - Ruano, Josue
AU - Arcila, John
AU - Romo-Bucheli, David
AU - Vargas, Carlos
AU - Rodríguez, Jefferson
AU - Mendoza, Oscar
AU - Plazas, Miguel
AU - Bautista, Lola
AU - Villamizar, Jorge
AU - Pedraza, Gabriel
AU - Moreno, Alejandra
AU - Valenzuela, Diana
AU - Vásquez, Lina
AU - Valenzuela-Santos, Carolina
AU - Camacho, Paúl
AU - Mantilla, Daniel
AU - Martínez, Fabio
N1 - Publisher Copyright:
© 2022,Biomedica. all rights reserved
PY - 2022
Y1 - 2022
N2 - Introduction: The coronavirus disease 2019 (COVID-19) has become a significant public health problem worldwide. In this context, CT-scan automatic analysis has emerged as a COVID-19 complementary diagnosis tool allowing for radiological finding characterization, patient categorization and disease follow-up. However, this analysis is dependent on the radiologist expertise, which might result in subjective evaluations. Objective: To explore deep learning representations, trained from thoracic CT-slices, to automatically distinguish COVID-19 disease from control samples. Materials and methods: Two datasets were used: SARS-CoV-2 CT Scan (Set-1) and FOSCAL dataset (Set-2). First, the deep representations take advantage of supervised learning models, previously trained on the natural image domain, which are adjusted following a transfer learning scheme. The deep classification was carried out: (a) via end-to-end deep learning approach and (b) via Random Forest and Support Vector Machine classifiers, by feeding the deep representation embedding vectors into these classifiers. Results: The End-to-end classification achieved an average accuracy of 92.33% (89.70% precision) for Set-1 and 96.99% (96.62% precision) for Set-2. The deep feature embedding with a Support Vector Machine achieved an average accuracy of 91.40% (95.77% precision) and 96.00% (94.74% precision), for Set-1 and Set-2 respectively. Conclusion: Deep representations have achieved outstanding performance in the identification of COVID-19 cases on CT Scans, demonstrating good characterization of the COVID-19 radiological patterns. These representations 5 could potentially support the COVID-19 diagnosis on clinical settings.
AB - Introduction: The coronavirus disease 2019 (COVID-19) has become a significant public health problem worldwide. In this context, CT-scan automatic analysis has emerged as a COVID-19 complementary diagnosis tool allowing for radiological finding characterization, patient categorization and disease follow-up. However, this analysis is dependent on the radiologist expertise, which might result in subjective evaluations. Objective: To explore deep learning representations, trained from thoracic CT-slices, to automatically distinguish COVID-19 disease from control samples. Materials and methods: Two datasets were used: SARS-CoV-2 CT Scan (Set-1) and FOSCAL dataset (Set-2). First, the deep representations take advantage of supervised learning models, previously trained on the natural image domain, which are adjusted following a transfer learning scheme. The deep classification was carried out: (a) via end-to-end deep learning approach and (b) via Random Forest and Support Vector Machine classifiers, by feeding the deep representation embedding vectors into these classifiers. Results: The End-to-end classification achieved an average accuracy of 92.33% (89.70% precision) for Set-1 and 96.99% (96.62% precision) for Set-2. The deep feature embedding with a Support Vector Machine achieved an average accuracy of 91.40% (95.77% precision) and 96.00% (94.74% precision), for Set-1 and Set-2 respectively. Conclusion: Deep representations have achieved outstanding performance in the identification of COVID-19 cases on CT Scans, demonstrating good characterization of the COVID-19 radiological patterns. These representations 5 could potentially support the COVID-19 diagnosis on clinical settings.
KW - Coronavirus infections/diagnosis
KW - Deep learning
KW - Tomography, x-ray computed
UR - http://www.scopus.com/inward/record.url?scp=85125100453&partnerID=8YFLogxK
U2 - 10.7705/biomedica.5927
DO - 10.7705/biomedica.5927
M3 - Artículo Científico
AN - SCOPUS:85125100453
SN - 0120-4157
VL - 42
JO - Biomedica
JF - Biomedica
IS - 1
ER -