A New Classification Method for Diagnosing COVID-19 Pneumonia via Joint Parallel Deformable MLP Modules and Bi-LSTM With Multi-Source Generated Data of CXR Images.

The Coronavirus Disease 2019 (COVID-19) pneumonia poses a critical threat to public health because of its powerful infectiousness, which has spurred the consumer electronics industry to innovate rapidly, leading to the emergence of advanced diagnostic and monitoring devices. Automated detection of COVID-19 based on multi-source generated data of CXR images is therefore essential for prevention. In the clinic, chest X-ray (CXR) plays a pivotal role in diagnosing COVID-19 pneumonia. However, it is challenging to utilize the dataset for diagnosis attributed to the significant imaging similarities observed among various types of pneumonia. To address the aforementioned issue, we present a pioneering classification model, denoted as PDMLP-Bi-LSTM, leveraging multi-source generated data. Its objective is to discriminate Normal, COVID-19, and Other pneumonia cases. Through the fusion of parallel deformable multi-layer perceptrons (MLPs) and Bi-directional Long Short-Term Memory (Bi-LSTM) modules, this model extracts multi-level abstract features and investigates potential correlations between parallel output features, capitalizing on the wealth of generated information. Initially, the chest region of the CXR image is localized and cropped using a pre-trained YOLO-V4 network, through which 13-dimensional transformed images and 16-dimensional depth feature maps are extracted using traditional image filters and convolutional neural network to form the 30-dimensional generated data for training the proposed classification model. The data is then fed spatially and channel-wise into deformable MLP modules, and the relationships of features on parallel channels are analyzed using Bi-LSTM modules. Finally, the classifier formed by fully connected layers and SoftMax function is employed to diagnose COVID-19 pneumonia. Extensive simulations based on 4099 CXR images were conducted to validate the performance of the proposed method. The results indicated that the proposed method exhibits excellent performance with accuracy, specificity, precision, recall, and F1-score by approximately 98% or above, which demonstrates the significant potential of the proposed method for clinically aiding in the diagnosis of patients with COVID-19 pneumonia.