Advances in artificial intelligence for accurate and timely diagnosis of COVID-19: A comprehensive review of medical imaging analysis

Abstract

In December 2019, the first case of coronavirus 2019 (COVID-19) appeared in China, quickly leading to a global pandemic. Early and accurate diagnosis is crucial for effective disease management. While reverse transcription polymerase chain reaction (RT-PCR) is the standard diagnostic test, it may yield false negative and misleading results. Artificial intelligence (AI) systems are accelerating the transformation of the medical field, particularly in early detection and diagnosis. Recent research has combined AI with medical imaging modalities, such as chest X-ray (CXR) and computed tomography (CT), to detect the virus, aiding doctors in making decisions and reducing misdiagnosis rates. In this article, we conducted a systematic review of high-quality articles published in high-impact journals that examined convolutional neural network (CNN)-based methods for detecting COVID-19 from radiographic or CT images and discussed associated issues. We synthesized publicly available datasets and evaluation measures, including accuracy, sensitivity, specificity, and F1 score, for each system used for automatic diagnosis of COVID-19 using several well-performing CNN architectures. Furthermore, we identified key research questions and future directions in this field. Our results show that the use of AI for COVID-19 diagnosis from radiographic and CT images has considerable potential to improve diagnostic accuracy and reduce false negative rates. Nevertheless, important challenges must be addressed, such as limited access to datasets and the need for rigorous model validation. Additionally, the generalization of models to different populations and contexts needs to be examined. Our findings underscore the need for future research directions, including the exploration of deep learning for smaller datasets, enhancing model performance for complex cases, and designing practical AI systems for deployment in clinical settings.

Ähnliche Arbeiten