The new coronavirus, known as COVID-19, first emerged in Wuhan, China, and since then has been transmitted to the whole world. Around 34 million people have been infected with COVID-19 virus so far, and nearly 1 million have died as a result of the virus. Resource shortages such as test kits and ventilator have arisen in many countries as the number of cases have increased beyond the control. Therefore, it has become very important to develop deep learning-based applications that automatically detect COVID-19 cases using chest X-ray images to assist specialists and radiologists in diagnosis. In this study, we propose a new approach based on deep LSTM model to automatically identify COVID-19 cases from X-ray images. Contrary to the transfer learning and deep feature extraction approaches, the deep LSTM model is an architecture, which is learned from scratch. Besides, the Sobel gradient and marker-controlled watershed segmentation operations are applied to raw images for increasing the performance of proposed model in the pre-processing stage. The experimental studies were performed on a combined public dataset constituted by gathering COVID-19, pneumonia and normal (healthy) chest X-ray images. The dataset was randomly separated into two sections as training and testing data. For training and testing, these separations were performed with the rates of 80%–20%, 70%–30% and 60%–40%, respectively. The best performance was achieved with 80% training and 20% testing rate. Moreover, the success rate was 100% for all performance criteria, which composed of accuracy, sensitivity, specificity and F-score. Consequently, the proposed model with pre-processing images ensured promising results on a small dataset compared to big data. Generally, the proposed model can significantly improve the present radiology based approaches and it can be very useful application for radiologists and specialists to help them in detection, quantity determination and tracing of COVID-19 cases throughout the pandemic.

Treatment of lung diseases, which are the third most common cause of death in the world, is of great importance in the medical field. Many studies using lung sounds recorded with stethoscope have been conducted in the literature in order to diagnose the lung diseases with artificial intelligence-compatible devices and to assist the experts in their diagnosis. In this paper, ICBHI 2017 database which includes different sample frequencies, noise and background sounds was used for the classification of lung sounds. The lung sound signals were initially converted to spectrogram images by using time-frequency method. The short time Fourier transform (STFT) method was considered as time-frequency transformation. Two deep learning based approaches were used for lung sound classification. In the first approach, a pre-trained deep convolutional neural networks (CNN) model was used for feature extraction and a support vector machine (SVM) classifier was used in classification of the lung sounds. In the second approach, the pre-trained deep CNN model was fine-tuned (transfer learning) via spectrogram images for lung sound classification. The accuracies of the proposed methods were tested by using the ten-fold cross validation. The accuracies for the first and second proposed methods were 65.5% and 63.09%, respectively. The obtained accuracies were then compared with some of the existing results and it was seen that obtained scores were better than the other results.

No abstract

The recognition of various lung sounds recorded using electronic stethoscopes plays a significant role in the early diagnoses of respiratory diseases. To increase the accuracy of specialist evaluations, machine learning techniques have been intensely employed during the past 30 years. In the current study, a new pretrained Convolutional Neural Network (CNN) model is proposed for the extraction of deep features. In the CNN architecture, an average-pooling layer and a max-pooling layer are connected in parallel in order to boost classification performance. The deep features are utilized as the input of the Linear Discriminant Analysis (LDA) classifier using the Random Subspace Ensembles (RSE) method. The proposed method was evaluated against a challenge dataset known as ICBHI 2017. The deep features and the LDA with RSE method provided the best accuracy score when compared to other existing methods using the same dataset, improving the classification accuracy by 5.75%. INDEX TERMS Lung sound, CNN model, parallel pooling, deep features, RSE method.

Cognitive prediction in the complicated and active environments is of great importance role in artificial learning. Classification accuracy of sound events has a robust relation with the feature extraction. In this paper, deep features are used in the environmental sound classification (ESC) problem. The deep features are extracted by using the fully connected layers of a newly developed Convolutional Neural Networks (CNN) model, which is trained in the end-to-end fashion with the spectrogram images. The feature vector is constituted with concatenating of the fully connected layers of the proposed CNN model. For testing the performance of the proposed method, the feature set is conveyed as input to the random subspaces K Nearest Neighbor (KNN) ensembles classifier. The experimental studies, which are carried out on the DCASE-2017 ASC and the UrbanSound8K datasets, show that the proposed CNN model achieves classification accuracies 96.23% and 86.70%, respectively. INDEX TERMS Environmental sound classification, spectrogram images, CNN model, deep features. DABAN ABDULSALAM ABDULLAH received the B.Sc. degree in computer science from the University of Human and Development, Iraq, in 2012, and the M.Sc. degree in applied mathematics and computer science from Eastern Mediterranean University, Cyprus, in 2015. He became a Lecturer with the Technical College of Informatics, Sulaimani Polytechnic University, in 2018. His research interests include data mining, association rules mining, and pattern recognition. ABDULKADIR SENGUR received the B.Sc. degree in electronics and computer education, the M.Sc. degree in electronics education, and the Ph.D. degree in electrical and electronics engineering from Firat University, Turkey, in 1999, 2003, and 2006, respectively. He became a Research Assistant with the Faculty of Technical Education, Firat University, in February 2001, where he is currently a Professor with the Faculty of Technical Education. His research interests include signal processing, image segmentation, pattern recognition, medical image processing, and computer vision.