Custom Deep Neural Network for 3D Covid Chest CT-scan Classification
This work addresses Covid-19 diagnosis through CT-scan classification, but it appears incremental as it builds on existing backbones without clear novel contributions.
The authors tackled the problem of classifying 3D chest CT-scans for Covid-19 diagnosis by proposing a custom deep neural network method that combines DenseNet 121 and ResNet 101 backbones, but no concrete results or numbers are provided in the abstract.
3D CT-scan base on chest is one of the controversial topisc of the researcher nowadays. There are many tasks to diagnose the disease through CT-scan images, include Covid19. In this paper, we propose a method that custom and combine Deep Neural Network to classify the series of 3D CT-scans chest images. In our methods, we experiment with 2 backbones is DenseNet 121 and ResNet 101. In this proposal, we separate the experiment into 2 tasks, one is for 2 backbones combination of ResNet and DenseNet, one is for DenseNet backbones combination.