COVID-19 CT Deep Learning Paper
Advancing Covidā19 differentiation with a robust preprocessing and integration of multiāinstitutional openārepository computer tomography datasets for deep learning analysis
I am pleased to announce that our second research article on COVID-19, Advancing Covidā19 differentiation with a robust preprocessing and integration of multiāinstitutional openārepository computer tomography datasets for deep learning analysis, has been published.
In this paper, we developed a Deep Learning pipeline for COVID-19 screening from CT images. Firstly, a stateāofātheāart custom UāNet model is applied to the CT image, in order to properly segment the lung areas. The segmentation model achieves a DICE similarity coefficient performance of 99.6%. Secondly, a classification network, which is based on a pre-trained VGG-19 model, is applied on the segmented images for COVIDā19 versus pneumonia differentiation and exhibits an area under curve (AUC) of 96.1%.
If you find it interesting, please cite it:
Trivizakis, E., Tsiknakis, N., Vassalou, E.E., Papadakis, G.Z., Spandidos, D.A., Sarigiannis, D., Tsatsakis, A., Papanikolaou, N., Karantanas, A.H. and Marias, K., 2020. Advancing COVIDā19 differentiation with a robust preprocessing and integration of multiāinstitutional openārepository computer tomography datasets for deep learning analysis. Experimental and Therapeutic Medicine. https://doi.org/10.3892/etm.2020.9210