2020/12/02 by Mario Manzo, Manzo, Mario, Simone Pellino +1
Computer Science · Medicine · #AI in cancer detection #COVID-19 diagnosis using AI #FOS: Electrical engineering #Image and Video Processing (eess.IV) #Radiomics and Machine Learning in Medical Imaging #electronic engineering #information engineering
paper · pdf · doi:10.48550/arxiv.2012.01251
openalex publication_date 2020/12/02 · openalex created_date 2022/07/25 · openalex updated_date 2026/07/28
The great challenge for the humanity of the year 2020 is the fight against\nCOVID-19. The whole world is making a huge effort to find an effective vaccine\nwith purpose to protect people not yet infected. The alternative solution\nremains early diagnosis, carried out through real-time polymerase chain\nreaction (RT-PCR) test or thorax computer tomography (CT) scan images. Deep\nlearning algorithms, specifically convolutional neural networks, represent a\nmethodology for the image analysis. They optimize the classification design\ntask, essential for an automatic approach on different types of images,\nincluding medical. In this paper, we adopt pretrained deep convolutional neural\nnetwork architectures in order to diagnose COVID-19 disease on CT images. Our\nidea is inspired by what the whole of humanity is achieving, substantially the\nset of multiple contributions is better than the single one for the fight\nagainst the pandemic. Firstly, we adapt, and subsequently retrain, for our\nassumption some neural architectures adopted in other application domains.\nSecondly, we combine the knowledge extracted from images by neural\narchitectures in an ensemble classification context. Experimental phase is\nperformed on CT images dataset and results obtained show the effectiveness of\nthe proposed approach with respect to state-of-the-art competitors.\n