A Systematic Survey on COVID 19 Detection and Diagnosis by Utilizing Deep Learning Techniques and Modalities of Radiology

Abstract

One of the most difficult aspects of the present COVID19 pandemic is early identification and diagnosis of COVID19, as well as exact segregation of non-COVID19 individuals at low cost and the sickness is in its early stages. Despite their widespread use in diagnostic centres, diagnostic approaches based solely on radiological imaging have flaws given the disease's novelty. As a result, to evaluate radiological pictures, healthcare practitioners and computer scientists frequently use machine learning and deep learning models. Based on a search strategy, from November 2019 to July 2020, researchers scanned the three different databases of Scopus, PubMed, and Web of Science for this study. Machine learning and deep learning are well-established artificial intelligence domains for data mining, analysis, and pattern recognition. Deep learning in which data is passed through many layers and automatically learning the composition of each layer from large dataset and it enables a new way that evaluates the complete image without human guidance to discern which insights are valuable, with applications ranging from object detection to medical image. Deep learning with CNN may have a significant effect on the automatic recognition and extraction of crucial features from X-ray and CT Scan images related to Covid19 analysis. According to the results, models based on deep learning possess amazing abilities to offer a precise and systematic system for detecting and diagnosing COVID19. In the field of COVID19 radiological imaging, deep learning software decreases false positive and false negative errors in the identification and diagnosis of the disease. It is providing a once-in-a-lifetime opportunity to provide patients with quick, inexpensive, and safe diagnostic services while also reducing the epidemic's impact on nursing and medical staff.

Ähnliche Arbeiten