Deep learning based fusion model for COVID-19 diagnosis and classification using computed tomography images.

Recently, the COVID-19 pandemic becomes increased in a drastic way, with the availability of a limited quantity of rapid testing kits. Therefore, automated COVID-19 diagnosis models are essential to identify the existence of disease from radiological images. Earlier studies have focused on the development of Artificial Intelligence (AI) techniques using X-ray images on COVID-19 diagnosis. This paper aims to develop a Deep Learning Based MultiModal Fusion technique called DLMMF for COVID-19 diagnosis and classification from Computed Tomography (CT) images. The proposed DLMMF model operates on three main processes namely Weiner Filtering (WF) based pre-processing, feature extraction and classification. The proposed model incorporates the fusion of deep features using VGG16 and Inception v4 models. Finally, Gaussian Naïve Bayes (GNB) based classifier is applied for identifying and classifying the test CT images into distinct class labels. The experimental validation of the DLMMF model takes place using open-source COVID-CT dataset, which comprises a total of 760 CT images. The experimental outcome defined the superior performance with the maximum sensitivity of 96.53%, specificity of 95.81%, accuracy of 96.81% and F-score of 96.73%.

Ectopic fat and aerobic fitness are key determinants of glucose homeostasis in nonobese Asians.

BACKGROUND: The importance of ectopic fat deposition and physical fitness in the pathogenesis of insulin resistance and beta cell dysfunction in subjects from the nonobese Asians is not known. MATERIALS AND METHODS: We conducted a cross-sectional study and measured insulin sensitivity (M value; 4-hour hyperinsulinaemic-euglycaemic clamp), insulin secretion rate (3-hour mixed meal tolerance test with oral minimal modelling), percent body fat, visceral adipose tissue, intramyocellular and intrahepatic lipid contents (magnetic resonance imaging and spectroscopy), cardiorespiratory fitness (VO2 max; graded exercise test) and habitual physical activity (short International Physical Activity Questionnaire) in 60 healthy nonobese Asian subjects (BMI = 21.9 ± 1.7 kg/m2 , age = 41.8 ± 13.4 years). RESULTS: M was inversely associated with percent body fat (r = -0.460, P < 0.001), visceral fat (r = -0.623, P < 0.001) and liver fat (r = -0.601, P < 0.001), whereas insulin secretion correlated positively with these adiposity indices (percent body fat: r = 0.303, P = 0.018; visceral fat: r = 0.409, P = 0.010; hepatic fat: r = 0.393, P = 0.002). VO2 max correlated negatively with insulin secretion rate (r = -0.420, P < 0.001) and positively with M (r = 0.658, P < 0.001). The amount of vigorous physical activity was positively associated with VO2 max (r = 0.682, P < 0.001). Multiple stepwise linear regression analyses indicated that VO2 max, age, and IHTG or VAT were independent determinants of insulin sensitivity and secretion (adjusted R2  = 69% and 33%, respectively, P < 0.001). CONCLUSIONS: Increased ectopic fat deposition is associated with reduced insulin sensitivity and increased insulin secretion in healthy nonobese Asians. Poor cardiorespiratory fitness, likely due to inadequate participation in vigorous exercise, is strongly related to suboptimal metabolic function. Interventions to encourage engagement in physical activity may thus be important for improving metabolic health in nonobese Asians.

Regulation of glucose metabolism in nondiabetic, metabolically obese normal-weight Asians.

Type 2 diabetes in Asia occurs largely in the absence of obesity. The metabolically obese normal-weight (MONW) phenotype refers to lean subjects with metabolic dysfunction that is typically observed in people with obesity and is associated with increased risk for diabetes. Previous studies evaluated MONW subjects who had greater body mass index (BMI) or total body fat than respective control groups, making interpretation of the results difficult. We evaluated insulin sensitivity (hyperinsulinemic-euglycemic clamp); insulin secretion (mixed meal with oral minimal modeling); intra-abdominal, muscle, and liver fat contents (magnetic resonance); and fasting and postprandial glucose and insulin concentrations in 18 MONW subjects and 18 metabolically healthy controls matched for age (43 ± 3 and 40 ± 3 yr; P = 0.52), BMI (both 22 ± 1 kg/m2; P = 0.69), total body fat (17 ± 1 and 16 ± 1 kg; P = 0.33), and sex (9 men and 9 women in each group). Compared with controls, MONW subjects had an approximately twofold greater visceral adipose tissue volume and an approximately fourfold greater intrahepatic fat content (but similar muscle fat), 20-30% lower glucose disposal rates and insulin sensitivity, and 30-40% greater insulin secretion rates (all P < 0.05). The disposition index, fasting glucose, and HbA1c concentrations were not different between groups, whereas postprandial glucose and insulin concentrations were ~15% and ~65% greater, respectively, in MONW than control subjects (both P < 0.05). We conclude that the MONW phenotype is associated with accumulation of fat in the intra-abdominal area and the liver, profound insulin resistance, but also a robust ß-cell insulin secretion response that compensates for insulin resistance and helps maintain glucose homeostasis.

Review on Deep Learning-Based CAD Systems for Breast Cancer Diagnosis.

Breast Cancer (BC) is a major health issue in women of the age group above 45. Identification of BC at an earlier stage is important to reduce the mortality rate. Image-based noninvasive methods are used for early detection and for providing appropriate treatment. Computer-Aided Diagnosis (CAD) schemes can support radiologists in making correct decisions. Computational intelligence paradigms such as Machine Learning (ML) and Deep Learning (DL) have been used in the recent past in CAD systems to accelerate diagnosis. ML techniques are feature driven and require a high amount of domain expertise. However, DL approaches make decisions directly from the image. The current advancement in DL approaches for early diagnosis of BC is the motivation behind this review. This article throws light on various types of CAD approaches used in BC detection and diagnosis. A survey on DL, Transfer Learning, and DL-based CAD approaches for the diagnosis of BC is presented in detail. A comparative study on techniques, datasets, and performance metrics used in state-of-the-art literature in BC diagnosis is also summarized. The proposed work provides a review of recent advancements in DL techniques for enhancing BC diagnosis.

Feature selection algorithm based on binary BAT algorithm and optimum path forest classifier for breast cancer detection using both echographic and elastographic mode ultrasound images.

Context: Breast cancer is one of the fatal diseases among women. Every year, its incidence and mortality rate increase globally. Mammography and sonography are widely used in breast cancer detection. Because mammography misses many cancers and shows false negatives in the denser tissues, sonography is preferred to give some extra information in addition to that available from mammography. Aims: To improve the performance of breast cancer detection by reducing false positives. Settings and Design: The local binary pattern (LBP) texture features must be extracted from ultrasound elastographic and echographic images of the same patients and then fused to form a single feature vector. Methods and Materials: Local Binary Pattern (LBP) texture features of elastographic and echographic images are extracted, and reduced individually through a hybrid feature selection technique based on binary BAT algorithm (BBA) and optimum path forest (OPF) classifier and then fused serially. Finally, the support vector machine classifier is used to classify the final fused feature set. Statistical Analysis Used: Various relevant performance metrics such as accuracy, sensitivity, specificity, discriminant power, Mathews correlation coefficient (MCC), F1 score, and Kappa were used to analyze the classification results. Results: The use of LBP feature produces 93.2% accuracy, 94.4% sensitivity, 92.3% specificity, 89.5% precision value, 91.88% F1 score, 93.34% balanced classification rate, and Mathews correlation coefficient of 0.861. The performance was compared with gray level co-occurrence matrix (GLCM), gray level difference matrix (GLDM), and LAWs features, and showed that LBP outperformed. Conclusions: Because the specificity is better, this method could be useful for detecting breast cancer with minimum false negatives.

Automated Brain Tumour Detection and Classification using Deep Features and Bayesian Optimised Classifiers.

PURPOSE: Brain tumour detection and classification require trained radiologists for efficient diagnosis. The proposed work aims to build a Computer Aided Diagnosis (CAD) tool to automate brain tumour detection using Machine Learning (ML) and Deep Learning (DL) techniques. MATERIALS AND METHODS: Magnetic Resonance Image (MRI) collected from the publicly available Kaggle dataset is used for brain tumour detection and classification. Deep features extracted from the global pooling layer of Pretrained Resnet18 network are classified using 3 different ML Classifiers, such as Support vector Machine (SVM), K-Nearest Neighbour (KNN), and Decision Tree (DT). The above classifiers are further hyperparameter optimised using Bayesian Algorithm (BA) to enhance the performance. Fusion of features extracted from shallow and deep layers of the pretrained Resnet18 network followed by BA-optimised ML classifiers is further used to enhance the detection and classification performance. The confusion matrix derived from the classifier model is used to evaluate the system's performance. Evaluation metrics, such as accuracy, sensitivity, specificity, precision, F1 score, Balance Classification Rate (BCR), Mathews Correlation Coefficient (MCC) and Kappa Coefficient (Kp), are calculated. RESULTS: Maximum accuracy, sensitivity, specificity, precision, F1 score, BCR, MCC, and Kp of 99.11 %, 98.99 %, 99.22 %, 99.09 %, 99.09 %, 99.10 %, 98.21 %, 98.21 %, respectively, were obtained for detection using fusion of shallow and deep features of Resnet18 pretrained network classified by BA optimized SVM classifier. Feature fusion performs better for classification task with accuracy, sensitivity, specificity, precision, F1 score, BCR, MCC and Kp of 97.31 %, 97.30 %, 98.65 %, 97.37 %, 97.34 %, 97.97%, 95.99 %, 93.95 %, respectively. CONCLUSION: The proposed brain tumour detection and classification framework using deep feature extraction from Resnet 18 pretrained network in conjunction with feature fusion and optimised ML classifiers can improve the system performance. Henceforth, the proposed work can be used as an assistive tool to aid the radiologist in automated brain tumour analysis and treatment.

Predicting the relationship between pesticide genotoxicity and breast cancer risk in South Indian women in in vitro and in vivo experiments.

Breast cancer is the third most common cancer in women after skin and lung cancer. Pesticides are of interest in etiologic studies of breast cancer because many pesticides mimic estrogen, a known breast cancer risk factor. In this study, we discerned the toxic role of the pesticides atrazine, dichlorvos, and endosulfan in inducing breast cancer. Various experimental studies, such as biochemical profiling of pesticide-exposed blood samples, comet assays, karyotyping analysis, pesticide and DNA interaction analysis by molecular docking, DNA cleavage, and cell viability assays, have been carried out. Biochemical profiling showed an increased level of blood sugar, WBC, hemoglobin, and blood urea in the patient exposed to pesticides for more than 15 years. The comet assay for DNA damage performed on patients exposed to pesticides and pesticide-treated blood samples revealed more DNA damage at the 50 ng concentration of all three pesticides. Karyotyping analysis showed enlargements in the heterochromatin region and 14pstk+, and 15pstk+in the exposed groups. In molecular docking analysis, atrazine had the highest glide score (- 5.936) and glide energy (- 28.690), which reveals relatively high binding capability with the DNA duplex. The DNA cleavage activity results showed that atrazine caused higher DNA cleavage than the other two pesticides. Cell viability was the lowest at 50 ng/ml (72 h). Statistical analysis performed using SPSS software unveiled a positive correlation (< 0.05) between pesticide exposure and breast cancer. Our findings support attempts to minimize pesticide exposure.

Particle swarm optimization based fusion of ultrasound echographic and elastographic texture features for improved breast cancer detection.

Breast cancer remains the main cause of cancer deaths among women in the world. As per the statistics, it is the most common killer disease of the new era. Since 2008, breast cancer incidences have increased by more than 20%, while mortality has increased by 14%. The statistics of breast cancer incidences as per GLOBOCAN project for the years 2008 and 2012 show an increase from 22.2 to 27% globally. In India, breast cancer accounts for 25% to 31% of all cancers in women. Mammography and Sonography are the two common imaging techniques used for the diagnosis and detection of breast cancer. Since Mammography fails to spot many cancers in the dense breast tissue of young patients, Sonography is preferred as an adjunct to Mammography to identify, characterize and localize breast lesions. This work aims to improve the performance of breast cancer detection by fusing the texture features from ultrasound elastographic and echographic images through Particle Swarm Optimization. The mean classification accuracy of Optimum Path Forest Classifier is used as an objective function in PSO. Seven performance metrics were computed to study the performance of the proposed technique using GLCM, GLDM, LAWs and LBP texture features through Support Vector Machine classifier. LBP feature provides accuracy, sensitivity, specificity, precision, F1 score, Mathews Correlation Coefficient and Balanced Classification Rate as 96.2%, 94.4%, 97.4%, 96.2%, 95.29%, 0.921, 95.88% respectively. The obtained performance using LBP feature is better compared to the other three features. An improvement of 6.18% in accuracy and 11.19% in specificity were achieved when compared to those obtained with previous works.

Fusion of k-Gabor features from medio-lateral-oblique and craniocaudal view mammograms for improved breast cancer diagnosis.

CONTEXT: Computer-aided diagnosis (CAD) combining mammographic features from cranio-caudal (CC) and medio-lateral-oblique (MLO) views improve the diagnostic performance of breast cancer. This could help doctors incorrect diagnosis at the earlier stage thereby reducing mortality. AIM: The aim of this study is to propose a breast cancer diagnostic technique for improving the diagnostic accuracy and reducing the false positive rate by fusing mammographic features from CC and MLO views. SETTINGS AND DESIGN: The MLO and CC view mammograms of same patients must be used to extract k-Gabor features and then fused to form a single feature vector. SUBJECTS AND METHODS: Mammograms from the digital database for screening mammography (DDSM) and INbreast datasets are collected. k-Gabor features extracted from both MLO and CC view mammograms are fused serially and reduced by principal component analysis (PCA) or genetic algorithm. The reduced features are classified using a multi-layer perceptron feed forward neural network with backpropagation learning algorithm. STATISTICAL ANALYSIS USED: Various relevant performance metrics such as accuracy, sensitivity, specificity, discriminant power, Mathews correlation coefficient (MCC), F1 score and Kappa are used to analyze the classification results. RESULTS: The accuracy, sensitivity, specificity, discriminant power, MCC, F1 score, and Kappa obtained as 92.5%, 93%, 91.8%, 1.198, 0.845, 0.936, and 0.845, respectively, for DDSM. For INbreast, the above specified metrics are 87.5%, 90.9%, 85.7%, 0.980, 0.741, 0.833, and 0.734, respectively. The results show 4.4%, 4.3%, and 9.4% improvements in accuracy, sensitivity, and specificity, respectively, compared to the previous works. CONCLUSIONS: Detailed analysis of the results implies that the serial fusion of k-Gabor features extracted from MLO and CC views with PCA reduction in CAD significantly improves the diagnostic performance.

Leptospirosis: clinical presentation and correlation with serovars.

A total of 2400 patients with pyrexia of unknown origin and or suspected leptospirosis were included in this study. Dark field microscopy detected Leptospira in 690 cases, Leptospira serological Investigations proved positive in 570 out of these 690 patients. Among them 212 had the classical icteric and the other 358 had anicteric type of presentation. Notably eptospira interrogans serovar ictero haemorrhagiae infection was encountered in 212 patients. In 30 patients, who had multi organ dysfunction which included renal failure, hepatic dysfunction or meningitis was due to Leptospira interrogans Serovar cannicola. Coexsistense of leptospirosis and hepatitis B virus infection were noted in 15 patients. Antibody to Leptospira interrogans was demonstrated by Micro agglutination test (MAT) in addition to dark field microscopy positivity in these cases. Similarly HIV antibody was demonstrated in 30 of the 330 anicteric patients. 554 out of 570 cases responded to intra venous penicillin (216), and oral Doxycycline (182) and Augmentin (156), and the remaining 16 patients succumbed to death.

In-vitro activity of polymyxin B, rifampicin, tigecycline alone and in combination against carbapenem-resistant Acinetobacter baumannii in Singapore.

OBJECTIVE: Carbapenem-resistant Acinetobacter baumannii (CR-AB) is an emerging cause of nosocomial infections worldwide. Combination therapy may be the only viable option until new antibiotics become available. The objective of this study is to identify potential antimicrobial combinations against CR-AB isolated from our local hospitals. METHODS: AB isolates from all public hospitals in Singapore were systematically collected between 2006 and 2007. MICs were determined according to CLSI guidelines. All CR-AB isolates were genotyped using a PCR-based method. Clonal relationship was elucidated. Time-kill studies (TKS) were conducted with polymyxin B, rifampicin and tigecycline alone and in combination using clinically relevant (achievable) unbound concentrations. RESULTS: 31 CR AB isolates were identified. They are multidrug-resistant, but are susceptible to polymyxin B. From clonal typing, 8 clonal groups were identified and 11 isolates exhibited clonal diversity. In single TKS, polymyxin B, rifampicin and tigecycline alone did not exhibit bactericidal activity at 24 hours. In combination TKS, polymyxin plus rifampicin, polymyxin B plus tigecycline and tigecycline plus rifampicin exhibited bactericidal killing in 13/31, 9/31 and 7/31 isolates respectively at 24 hours. Within a clonal group, there may be no consensus with the types of antibiotics combinations that could still kill effectively. CONCLUSION: Monotherapy with polymyxin B may not be adequate against polymyxin B susceptible AB isolates. These findings demonstrate that in-vitro synergy of antibiotic combinations in CR AB may be strain dependant. It may guide us in choosing a pre-emptive therapy for CR AB infections and warrants further investigations.

Impact of harvesting on constructed wetlands performance - a comparison between Scirpus grossus and Typha angustifolia.

Three units of free water surface (FWS) constructed wetlands treating domestic wastewater under tropical conditions were examined in terms of water quality and biomass characteristics. One unit (L2) was planted with Scirpus grossus, one with Typha angustifolia (L3), and the unplanted third (L1) served as control. Influent and effluent quality parameters: biological oxygen demand (BOD(5)), nitrate (NO(3)(-)-N), ammonium (NH(4)(+)-N), phosphorus (P), total suspended solids (TSS) and fecal coliforms were regularly measured. The average BOD(5) reductions were 37.0%, 58.5%, and 53.8% for units L1, L2, and L3, respectively. The planted units removed pollutants more effectively although there was no significant difference between the Scirpus grossus and Typha angustifolia units. Plant growth was monitored in marked quadrats by measuring shoot height and other growth parameters. The above-ground organs in L2 and L3 was harvested whenever the shoots reached maximum shoot height and formed flowers. Scirpus grossus had sustainable above-ground biomass production but Typha angustifolia could not sustain repeated harvestings with the above-ground biomass production declining significantly following four consecutive harvests.

Fusarium solani breast abscess.

An unusual manifestation of breast fusariosis was encountered in a 55-year-old female diabetic patient. Two fine needle aspirates (FNA) from the abscess were done at three days interval and they showed hyaline, septate, branched, fungal hyphae in 10% potassium hydroxide mount. Fungal infection was confirmed by demonstrating the fungal hyphae in the midst of lymphocytes, macrophages and neutrophils in Leishman stained smears. Culture of both FNAs yielded a heavy and pure growth of Fusarium solani. The patient responded to oral ketoconazole 200 mg once daily for 3 weeks. The breast fusariosis reported here is presumably the first case in India.