COVID-19 Detection from X-rays using CNN-based Graph with Fast Localization spectral filters

The global impact of the COVID-19 pandemic has been severe, causing extensive disruption and damage. It's been going on for approximately four years, and despite the development of vaccines, new viral varieties continue to appear and spread. Early detection of COVID-19 is critical for controlling its spread, and chest X-rays can help. On the other hand, deep learning-based algorithms for detecting different characteristics in chest X-rays can take a long time to train, which can be significant in the event of fast-spreading novel variants. This paper offers a COVID-19 chest X-ray classification system based on Graph with Fast Localization spectral filters (GFLs) to solve this issue. The CNN-GFL model demonstrates a rapid training capability, enabling accelerated and highly accurate COVID-19 predictions. The effectiveness of the CNN-GFL model in COVID-19 diagnosis was evaluated by conducting experiments on a publicly available dataset of COVID-19 chest X-rays. The obtained results demonstrated the commendable performance of the model in accurately detecting COVID-19 cases. The study emphasises the promise of deep learning-based methods for COVID-19 diagnosis and the significance of faster and more accurate diagnosis to restrict the virus's spread. The proposed method may be helpful in cases when new viral varieties develop and disseminate quickly, allowing for early diagnosis and reaction to outbreaks. The study also highlights the need for continued research to manage the pandemic effectively. Overall, the proposed approach could be a valuable tool in the fight against COVID-19, enabling faster and more accurate diagnosis to contain its spread.