Using neural networks for classification of COVID-19 cases

Neural nets are three-dimensional meshes of neurons modeled after the brain. They are also called "connectionist" models because they are a vast collection of randomly interconnected elements. Attempts to program neural nets oftentimes yield limitless and unpredictable outcomes that seem to display intelligence. In this paper, the author use neural networks for the classification of COVID-19 cases. Neural networks are biologically inspired, with multiple layers that are densely interconnected units. The author compare COVID-19 cases by looking into CT scan reports. CT scan reports are interpreted by reading a few images and matching with the three-dimensional medical scans. Incidentally, these images have to be read with no change in the healthcare and diagnostics industries. The findings indicate that DenseNet-121 renders the best results for this classification problem.