Table 1 Recent related works.

⁸2020

Proposing a triple approach aimed at differentiating between general lung disease and COVID-19 and highlighting symptomatic areas of COVID-19 disease on chest radiographs when COVID-19 disease is detected

Only significant differences between COVID-19 and other pulmonary diseases were studied

⁹2021

A method is proposed to resize the image using the maximum window function, which preserves the anatomical structure of the chest

Only network models after using transfer learning were studied

¹⁰2021

The proposed deep learning model based on ResNet 50, named CORNet, was used to detect COVID-19, and a retrospective multicenter analysis was also performed to extract visual features from volumetric chest CT scans during COVID-19 detection

The use of traditional network convergence strategies results in poor accuracy metrics

¹¹2021

A new framework of cascaded deep learning classifiers enhances the performance of CAD systems for suspected COVID-19 and pneumonia diseases in X-ray images

It was composed of conventional networks, resulting in too many network parameters

¹²2023

tested several CNNs, focusing on the three architectures with the best performance, namely InceptionV3, DenseNet201, and EfficientNetB3

Deep learning classification of captured pulmonary sounds, but with low accuracy due to noise

¹³2023

The 2D CNN model was used for optimal feature extraction with minimal time and space requirements. The CNN features were then classified using various machine-learning classifiers

The classical machine learning algorithm is used to extract features without considering the limitations of a single model