Bioinspired CNN Approach for Diagnosing COVID-19 Using Images of Chest X-Ray

ABSTRACT

Introduction: The provision of medical facilities needed for COVID-19 diagnosis is a global concern. They must be a powerful tool for detecting and diagnosing the virus quickly using a variety of tests, as well as low-cost advancements. Whereas a chest X-ray image is an effective screening technique, the image acquisition by the instruments must be read appropriately and quickly if multiple tests are performed. Objectives: COVID-19 causes continuous respiratory parenchymal ground glass and integrates respiratory opacity, with a curved shape and peripheral pulmonary dissemination in some cases, which is difficult to anticipate earlier on. In this chapter, we intend to construct a good platform to identify COVID-19 characteristics from the image of chest X-ray to aid in early analysis. Methods: In particular, based on the Cuckoo search method, this chapter provides a bioinspired CNN-based model for COVID-19 diagnosis. This method identifies different deep learning strategies of COVID-19 patients' chest X-ray images for accurate infection identification. The suggested model's performance is estimated using the Cuckoo search approach. Furthermore, the bioinspired CNN characteristics are fine-tuned using optimization algorithm. Rigorous testing reveals that suggested method may accurately categorize chest X-ray images with high performance, remembrance, and sensitivity. Results: As a result, the suggested approach can be used to classify COVID-19 diseases from chest X-ray images in real time and also accuracy will be validated. Conclusion: Finally, the investigation of comparison results illustrates the Cuckoo algorithm is realized to determine the interested regions of the COVID-19 x-ray images. © 2023, The Author(s), under exclusive license to Springer Nature Switzerland AG.