COVID­19 detection from chest X-ray images using transfer learning.

COVID-19 is a kind of coronavirus that appeared in China in the Province of Wuhan in December 2019. The most significant influence of this virus is its very highly contagious characteristic which may lead to death. The standard diagnosis of COVID-19 is based on swabs from the throat and nose, their sensitivity is not high enough and so they are prone to errors. Early diagnosis of COVID-19 disease is important to provide the chance of quick isolation of the suspected cases and to decrease the opportunity of infection in healthy people. In this research, a framework for chest X-ray image classification tasks based on deep learning is proposed to help in early diagnosis of COVID-19. The proposed framework contains two phases which are the pre-processing phase and classification phase which uses pre-trained convolution neural network models based on transfer learning. In the pre-processing phase, different image enhancements have been applied to full and segmented X-ray images to improve the classification performance of the CNN models. Two CNN pre-trained models have been used for classification which are VGG19 and EfficientNetB0. From experimental results, the best model achieved a sensitivity of 0.96, specificity of 0.94, precision of 0.9412, F1 score of 0.9505 and accuracy of 0.95 using enhanced full X-ray images for binary classification of chest X-ray images into COVID-19 or normal with VGG19. The proposed framework is promising and achieved a classification accuracy of 0.935 for 4-class classification.

Framework of Computer Aided Diagnosis Systems for Cancer Classification Based on Medical Images.

Early detection of cancer can increase patients' survivability and treatment options. Medical images such as Mammogram, Ultrasound, Magnetic Resonance Imaging, and microscopic images are the common method for cancer diagnosis. Recently, computer-aided diagnosis (CAD) systems have been used to help physicians in cancer diagnosis so that the diagnosis accuracy can be improved. CAD can help in decreasing missed cancer lesions due to physician fatigue, reducing the burden of workload and data overloading, and decreasing variability of inter- and intra-readers of images. In this research, a framework of CAD systems for cancer diagnosis based on medical images has been proposed. The proposed work helps physicians in detection of suspicion regions using different medical images modalities and in classifying the detected suspicious regions as normal or abnormal with the highest possible accuracy. The proposed framework of CAD system consists of four stages which are: preprocessing, segmentation of regions of interest, feature extraction and selection, and finally classification. In this research, the framework has been applied on blood smear images to diagnose the cases as normal or abnormal for Acute Lymphoblastic Leukemia (ALL) cases. Ant Colony Optimization (ACO) has been used to select the subsets of features from the features extracted from segmented cell parts which can maximize the classification performance as possible. Different classifiers which are Decision Tree (DT), K-nearest neighbor (K-NN), Naïve Bayes (NB), and Support Vector Machine (SVM) have been applied. The framework has been yielding promising results which reached 96.25% accuracy, 97.3% sensitivity, and 95.35% specificity using decision tree classifier.

Machine learning techniques for breast cancer computer aided diagnosis using different image modalities: A systematic review.

BACKGROUND AND OBJECTIVE: The high incidence of breast cancer in women has increased significantly in the recent years. Physician experience of diagnosing and detecting breast cancer can be assisted by using some computerized features extraction and classification algorithms. This paper presents the conduction and results of a systematic review (SR) that aims to investigate the state of the art regarding the computer aided diagnosis/detection (CAD) systems for breast cancer. METHODS: The SR was conducted using a comprehensive selection of scientific databases as reference sources, allowing access to diverse publications in the field. The scientific databases used are Springer Link (SL), Science Direct (SD), IEEE Xplore Digital Library, and PubMed. Inclusion and exclusion criteria were defined and applied to each retrieved work to select those of interest. From 320 studies retrieved, 154 studies were included. However, the scope of this research is limited to scientific and academic works and excludes commercial interests. RESULTS: This survey provides a general analysis of the current status of CAD systems according to the used image modalities and the machine learning based classifiers. Potential research studies have been discussed to create a more objective and efficient CAD systems.

Semantic annotation for biological information retrieval system.

Online literatures are increasing in a tremendous rate. Biological domain is one of the fast growing domains. Biological researchers face a problem finding what they are searching for effectively and efficiently. The aim of this research is to find documents that contain any combination of biological process and/or molecular function and/or cellular component. This research proposes a framework that helps researchers to retrieve meaningful documents related to their asserted terms based on gene ontology (GO). The system utilizes GO by semantically decomposing it into three subontologies (cellular component, biological process, and molecular function). Researcher has the flexibility to choose searching terms from any combination of the three subontologies. Document annotation is taking a place in this research to create an index of biological terms in documents to speed the searching process. Query expansion is used to infer semantically related terms to asserted terms. It increases the search meaningful results using the term synonyms and term relationships. The system uses a ranking method to order the retrieved documents based on the ranking weights. The proposed system achieves researchers' needs to find documents that fit the asserted terms semantically.

A framework for prediction of response to HCV therapy using different data mining techniques.

Hepatitis C which is a widely spread disease all over the world is a fatal liver disease caused by Hepatitis C Virus (HCV). The only approved therapy is interferon plus ribavirin. The number of responders to this treatment is low, while its cost is high and side effects are undesirable. Treatment response prediction will help in reducing the patients who suffer from the side effects and high costs without achieving recovery. The aim of this research is to develop a framework which can select the best model to predict HCV patients' response to the treatment of HCV from clinical information. The framework contains three phases which are preprocessing phase to prepare the data for applying Data Mining (DM) techniques, DM phase to apply different DM techniques, and evaluation phase to evaluate and compare the performance of the built models and select the best model as the recommended one. Different DM techniques had been applied which are associative classification, artificial neural network, and decision tree to evaluate the framework. The experimental results showed the effectiveness of the framework in selecting the best model which is the model built by associative classification using histology activity index, fibrosis stage, and alanine amino transferase.