A New Model to Detect COVID-19 Coughing and Breathing Sound Symptoms Classification from CQT and Mel Spectrogram Image Representation using Deep Learning

Deep Learning is a relatively new Artificial Intelligence technique that has shown to be extremely effective in a variety of fields. Image categorization and also the identification of artefacts in images are being employed in visual recognition. The goal of this study is to recognize COVID-19 artefacts like cough and also breath noises in signals from realworld situations. The suggested strategy considers two major steps. The first step is a signal-to-image translation that is aided by the Constant-Q Transform (CQT) and a Mel-scale spectrogram method. Next, nine deep transfer models (GoogleNet, ResNet18/34/50/100/101, SqueezeNet, MobileNetv2, and NasNetmobile) are used to extract and also categorise features. The digital audio signal will be represented by the recorded voice. The CQT will transform a time-domain audio input to a frequency-domain signal. To produce a spectrogram, the frequency will really be converted to a log scale as well as the colour dimension will be converted to decibels. To construct a Mel spectrogram, the spectrogram will indeed be translated onto a Mel scale. The dataset contains information from over 1,600 people from all over the world (1185 men as well as 415 women). The suggested DL model takes as input the CQT as well as Melscale spectrograms derived from the breathing and coughing tones of patients diagnosed using the coswara-combined dataset. With the better classification performance employing cough sound CQT and a Mel-spectrogram image, the current proposal outperformed the other nine CNN networks. For patients diagnosed, the accuracy, sensitivity, as well as specificity were 98.9%, 97.3%, and 98.1%, respectively. The Resnet18 is the most reliable network for symptomatic patients using cough and breath sounds. When applied to the Coswara dataset, we discovered that the suggested model's accuracy (98.7%) outperforms the state-of-the-art models (85.6%, 72.9%, 87.1%, and 91.4%) according to the SGDM optimizer. Finally, the method as a main screening tool to try and identify COVID-19 by of disease transmission.

A new method for detecting COVID-19 combining deep learning with improved histogram equalization and median filter

The new coronavirus (COVID-2019) epidemic outbreak has devastating impacts on people's daily lives and public healthcare systems. The chest X-ray image is an effective tool for diagnosing new coronavirus diseases. This paper proposes a new method to identify the new coronavirus from chest X-ray images to assist radiologists in fast and accurate image reading. We first enhance the contrast of X-ray images by using adaptive histogram equalization and eliminating image noise by using a median filter. Then, the X-ray image is fed to a sophisticated deep neural network (FAC-DPN-SENet) proposed by us to train a classifier, which is used to classify an X-ray image as usual or COVID-2019 or other pneumonia. Applying our method to a standard dataset, we achieve a classification accuracy of 93%, which is significantly better performance than several other state-of-the-art models, such as ResNet and DenseNet. This shows that the proposed method can be used as an effective tool to detect COVID-2019. © 2021 IEEE.

Covid-19 Classification Using HOG-SVM and Deep Learning Models

COVID-19 is measured as the biggest hazardous and fast infectious grief for the human body which has a severe impact on lives, health, and the community all over the world. It is still spreading throughout the world with different variants which is silently killing many lives globally. Thus, earlier diagnosis and accurate detection of COVID-19 cases are essential to protect global lives. Diagnosis COVID-19 through chest X-ray images is one of the best solutions to detect the virus in the infected person properly and quickly at a low cost. Encouraged by the existing research, in this paper, we proposed a hybrid model to classify the Covid cases and non-Covid cases with chest X-ray images based on feature extraction, machine learning and deep learning techniques. Two feature extractors, Histogram Oriented Gradient (HOG) and CNN (MobileNetV2, Sequential, ResNet152V2) are used to train the model. For the classification, we utilized two approaches: Support Vector Machine (SVM) for machine learning and CNN (MobileNetV2, Sequential, ResNet152V2) classifiers for deep learning. The experimental result analysis shows that the Sequential model and the ResNet152V2 model achieve 100% and 82.6% accuracy respectively which is satisfactory. On the other hand, the HOG-SVM method successfully detects all the test images correctly which provides the best result with 100% accuracy, specificity, and responsiveness over a limited public dataset. © 2022 IEEE.

COVID-19 CT image denoising algorithm based on adaptive threshold and optimized weighted median filter.

CT image of COVID-19 is disturbed by impulse noise during transmission and acquisition. Aiming at the problem that the early lesions of COVID-19 are not obvious and the density is low, which is easy to confuse with noise. A median filtering algorithm based on adaptive two-stage threshold is proposed to improve the accuracy for noise detection. In the advanced stage of ground-glass lesion, the density is uneven and the boundary is unclear. It has similar gray value to the CT images of suspected COVID-19 cases such as adenovirus pneumonia and mycoplasma pneumonia (reticular shadow and strip shadow). Aiming at the problem that the traditional weighted median filter has low contrast and fuzzy boundary, an adaptive weighted median filter image denoising method based on hybrid genetic algorithm is proposed. The weighted denoising parameters can adaptively change according to the detailed information of lung lobes and ground-glass lesions, and it can adaptively match the cross and mutation probability of genetic combined with the steady-state regional population density, so as to obtain a more accurate COVID-19 denoised image with relatively few iterations. The simulation results show that the improved algorithm under different density of impulse noise is significantly better than other algorithms in peak signal-to-noise ratio (PSNR), image enhancement factor (IEF) and mean absolute error (MSE). While protecting the details of lesions, it enhances the ability of image denoising.