An externally validated fully automated deep learning algorithm to classify COVID-19 and other pneumonias on chest computed tomography

Abstract

PurposeIn this study, we propose an Artificial Intelligence framework based on 3D Convolutional Neural network (CNN) to classify CT scans of patients with COVID-19, Influenza/CAP, and no-infection, after automatic segmentation of the lungs and lung abnormalities.MethodsThe AI classification model is based on inflated 3D Inception architecture and was trained and validated on retrospective data of CT images of 667 adult patients (No infection: 188, COVID-19: 230, Influenza/CAP: 249) and 210 adult patients (No infection: 70, COVID-19: 70, Influenza/CAP: 70), respectively. The model's performance was independently evaluated on an internal test set of 273 adult patients (No infection: 55, COVID-19: 94, Influenza/CAP: 124) and an external validation set coming from a different center (305 adult patients, COVID-19: 169, No infection: 76, Influenza/CAP: 60).ResultsThe model showed excellent performance in the external validation set with an AUC of 0.90, 0.92 and 0.92 for COVID-19, Influenza/CAP and No infection respectively. The selection of the input slices based on automatic segmentation of the abnormalities in the lung reduces analysis time (56 s per scan) and computational burden of the model. The TRIPOD score of the proposed model is 47% (15 out of 32 TRIPOD items).ConclusionThis AI solution provides rapid and accurate diagnosis in patients suspected of COVID-19 infection and influenza.

ACSN: Attention capsule sampling network for diagnosing COVID-19 based on chest CT scans

“…Clearly, the proposed ACSN model showed superior performance compared to other methods in all indicators compared with the results published by Covid-Fact [ 40 ] and Vaidyanathan [ 41 ]. Moreover, compared with the results of Mohamed et al’s study [ 43 ], the ACSN model was not superior in all indicators but did surpass it in some regards, especially in terms of specificity, with 1.4% better results.…”

Inspection of chicken wings and legs for animal welfare monitoring using X-ray computed tomography, visual examination, and histopathology

“…For instance, machine learning algorithms can be used to analyze images obtained through X-ray and/or visible light imaging, thereby enabling the automatic classification of limb injuries cases by the trained model. In the literature, numerous studies have investigated the use of machine learning in interpreting CT scans for human diagnostics ( Lassau et al, 2021 ; Vaidyanathan et al, 2022 ). For example, predicting the severity of COVID-19 patients based on chest CT scans is one such application.…”

The Relationship Between OR-PNI Score and Disease Course in Coronavirus-19 Infection

“…[22] Algorithms and artificial intelligence models created using COVID-19 thoracic CT imaging data provide valuable insights into the diagnosis and prognosis of the disease. [23,24] The COVID-19 Reporting and Data System were established in 2020 with the aim of standardizing COVID-19 CT findings. [25] Microthrombi and the coagulation cascade are important in the pathophysiology of COVID-19 infection.…”

“…FARI has been presented as a marker of thrombosis risk, inflammation, and nutritional status. [23] In the present study, we compared OR-PNI with FARI and PNI, as they are considered predictors of mortality in intense inflammatory processes such as COVID-19. The results showed that the FARI value is effective in predicting mortality rather than PNI and OR-PNI.…”