Computer-Aided Detection of COVID-19 from CT Images Based on Gaussian Mixture Model and Kernel Support Vector Machines Classifier

Automated Screening of COVID-19-Based Tongue Image on Chinese Medicine

Objective

Artificial intelligence-powered screening systems of coronavirus disease 2019 (COVID-19) are urgently demanding since the ongoing outbreak of SARS-CoV-2 worldwide. Chest CT or X-ray is not sufficient to support the large-scale screening of COVID-19 because mildly-infected patients do not have imaging features on these images. Therefore, it is imperative to exploit supplementary medical imaging strategies. Traditional Chinese medicine has played an essential role in the fight against COVID-19.

Methods

In this paper, we conduct two kinds of verification experiments based on a newly-collected multi-modality dataset, which consists of three types of modalities: tongue images, chest CT scans, and X-ray images. First, we study a binary classification experiment on tongue images to verify the discriminative ability between COVID-19 and non-COVID-19. Second, we design extensive multimodality experiments to validate whether introducing tongue image can improve the screening accuracy of COVID-19 based on chest CT or X-ray images.

Results

Tongue image screening of COVID-19 showed that the accuracy (ACC), sensitivity (SEN), specificity (SPEC), and Matthew correlation coefficient (MCC) of the improved AlexNet and Googlenet both reached 98.39%, 98.97%, 96.67%, and 99.11%. The fusion of chest CT and tongue images used a tandem multimodal classifier fusion strategy to achieve optimal classification, and the results and screening accuracy of COVID-19 reached 98.98%, resulting in a significant improvement of 4.75% the highest accuracy in 375 years compared with the single-modality model. The fusion of chest x-rays and tongue images also had good classification accuracy.

Conclusions

Both experimental results demonstrate that tongue image not only has an excellent discriminative ability for screening COVID-19 but also can improve the screening accuracy based on chest CT or X-rays. To the best of our knowledge, it is the first work that verifies the effectiveness of tongue image on screening COVID-19. This paper provides a new perspective and a novel solution that contributes to large-scale screening toward fast stopping the pandemic of COVID-19.

Machine learning based COVID -19 disease recognition using CT images of SIRM database

The COVID-19 pandemic, probably one of the most widespread pandemics humanity has encountered in the twenty first century, caused death to almost 1.75 M people worldwide, impacting almost 80 M lives with direct contact. In order to contain the spread of coronavirus, it is necessary to develop a reliant and quick method to identify those who are affected and isolate them until full recovery is made. The imagery knowledge has been shown to be useful for quick COVID-19 diagnosis. Though the scans of computational tomography (CT) demonstrate a range of viral infection signals, considering the vast number of images, certain visual characteristics are challenging to distinguish and can take a long time to be identified by radiologists. In this study for detection of the COVID-19, a dataset is formed by taking 3764 images. The feature extraction process is applied to the dataset to increase the classification performance. Techniques like Grey Level Co-occurrence Matrix (GLCM) and Discrete Wavelet Transform (DWT) are used for feature extraction. Then various machine learning algorithms applied such as Support Vector Machines (SVM), Linear Discriminant Analysis (LDA), Multi- Level Perceptron, Naive Bayes, K-Nearest Neighbours and Random Forests are used for classification of COVID-19 disease detection. Sensitivity, Specificity, Accuracy, Precision, and F-score are the metrics used to measure the performance of different machine learning models. Among these machine learning models SVM with GLCM as feature extraction technique using 10-fold cross validation gives the best classification result with 99.70% accuracy, 99.80% sensitivity and 97.03% F-score. We also ran these tests on different data sets and found that the results are similar across those too, as discussed later in the results section.

COVID-19 Diagnosis from Medical Images Using Transfer Learning

Introduction

The novel coronavirus (COVID-19) originated in Wuhan, China, in December 2019. To date, the virus has infected more than 110 million people worldwide and claimed 2.5 million lives. With the rapid increase in the number of infected cases, some countries face a shortage of testing resources. Computational methods such as deep learning algorithms can help in such a situation to expedite and automate the diagnosis of COVID-19.

Methods

In this research, we trained eight convolutional neural network models to automatically detect and diagnose COVID-19 from medical imaging, including X-ray and CT scan images. Those deep learning networks have a predefined structure in which we re-train on medical images to serve our purpose, which is called transfer learning.

Results

We used two different medical images known as X-ray and CT scan. The experimental results show that CT scan achieved better performance than X-ray. Specifically, the Xception network model has achieved an overall performance on CT scan of 84%, 91%, and 77% for accuracy, sensitivity, and specificity, respectively. That was the highest in all models that we trained. On the other hand, the same network model (Xception) was applied on X-ray and performed 69%, 83%, and 55% for accuracy, sensitivity, and specificity, respectively.

Conclusion

The performance of our proposed model to detect COVID-19 from CT scan is acceptable and promising to start in the field. We target the medical sectors to help them by providing rapid and accurate diagnosis of COVID-19 cases using an alternative detection approach to the traditional ones.