UDBRNet: A novel uncertainty driven boundary refined network for organ at risk segmentation.

Organ segmentation has become a preliminary task for computer-aided intervention, diagnosis, radiation therapy, and critical robotic surgery. Automatic organ segmentation from medical images is a challenging task due to the inconsistent shape and size of different organs. Besides this, low contrast at the edges of organs due to similar types of tissue confuses the network's ability to segment the contour of organs properly. In this paper, we propose a novel convolution neural network based uncertainty-driven boundary-refined segmentation network (UDBRNet) that segments the organs from CT images. The CT images are segmented first and produce multiple segmentation masks from multi-line segmentation decoder. Uncertain regions are identified from multiple masks and the boundaries of the organs are refined based on uncertainty data. Our method achieves remarkable performance, boasting dice accuracies of 0.80, 0.95, 0.92, and 0.94 for Esophagus, Heart, Trachea, and Aorta respectively on the SegThor dataset, and 0.71, 0.89, 0.85, 0.97, and 0.97 for Esophagus, Spinal Cord, Heart, Left-Lung, and Right-Lung respectively on the LCTSC dataset. These results demonstrate the superiority of our uncertainty-driven boundary refinement technique over state-of-the-art segmentation networks such as UNet, Attention UNet, FC-denseNet, BASNet, UNet++, R2UNet, TransUNet, and DS-TransUNet. UDBRNet presents a promising network for more precise organ segmentation, particularly in challenging, uncertain conditions. The source code of our proposed method will be available at https://github.com/riadhassan/UDBRNet.

RVCNet: A hybrid deep neural network framework for the diagnosis of lung diseases.

Early evaluation and diagnosis can significantly reduce the life-threatening nature of lung diseases. Computer-aided diagnostic systems (CADs) can help radiologists make more precise diagnoses and reduce misinterpretations in lung disease diagnosis. Existing literature indicates that more research is needed to correctly classify lung diseases in the presence of multiple classes for different radiographic imaging datasets. As a result, this paper proposes RVCNet, a hybrid deep neural network framework for predicting lung diseases from an X-ray dataset of multiple classes. This framework is developed based on the ideas of three deep learning techniques: ResNet101V2, VGG19, and a basic CNN model. In the feature extraction phase of this new hybrid architecture, hyperparameter fine-tuning is used. Additional layers, such as batch normalization, dropout, and a few dense layers, are applied in the classification phase. The proposed method is applied to a dataset of COVID-19, non-COVID lung infections, viral pneumonia, and normal patients' X-ray images. The experiments take into account 2262 training and 252 testing images. Results show that with the Nadam optimizer, the proposed algorithm has an overall classification accuracy, AUC, precision, recall, and F1-score of 91.27%, 92.31%, 90.48%, 98.30%, and 94.23%, respectively. Finally, these results are compared with some recent deep-learning models. For this four-class dataset, the proposed RVCNet has a classification accuracy of 91.27%, which is better than ResNet101V2, VGG19, VGG19 over CNN, and other stand-alone models. Finally, the application of the GRAD-CAM approach clearly interprets the classification of images by the RVCNet framework.

A novel approach for standardizing clinical laboratory categorical test results using machine learning and string distance similarity.

Standardizing clinical laboratory test results is critical for conducting clinical data science research and analysis. However, standardized data processing tools and guidelines are inadequate. In this paper, a novel approach for standardizing categorical test results based on supervised machine learning and the Jaro-Winkler similarity algorithm is proposed. A supervised machine learning model is used in this approach for scalable categorization of the test results into predefined groups or clusters, while Jaro-Winkler similarity is used to map text terms into standard clinical terms within these corresponding groups. The proposed method is applied to 75062 test results from two private hospitals in Bangladesh. The Support Vector Classification algorithm with a linear kernel has a classification accuracy of 98%, which is better than the Random Forest algorithm when categorizing test results. The experiment results show that Jaro-Winkler similarity achieves a remarkable 99.93% success rate in the test result standardization for the majority of groups with manual validation. The proposed method outperforms previous studies that concentrated on standardizing test results using rule-based classifiers on a smaller number of groups and distance similarities such as Cosine similarity or Levenshtein distance. Furthermore, when applied to the publicly available MIMIC-III dataset, our approach also performs excellently. All these findings show that the proposed standardization technique can be very beneficial for clinical big data research, particularly for national clinical research data hubs in low- and middle-income countries.

Quantum dynamics simulations of the 2D spectroscopy for exciton polaritons.

We develop an accurate and numerically efficient non-adiabatic path-integral approach to simulate the non-linear spectroscopy of exciton-polariton systems. This approach is based on the partial linearized density matrix approach to model the exciton dynamics with explicit propagation of the phonon bath environment, combined with a stochastic Lindblad dynamics approach to model the cavity loss dynamics. Through simulating both linear and polariton two-dimensional electronic spectra, we systematically investigate how light-matter coupling strength and cavity loss rate influence the optical response signal. Our results confirm the polaron decoupling effect, which is the reduced exciton-phonon coupling among polariton states due to the strong light-matter interactions. We further demonstrate that the polariton coherence time can be significantly prolonged compared to the electronic coherence outside the cavity.

Wide variation of heterozygotic genotypes of recent fasciolid hybrids from livestock in Bangladesh assessed by rDNA internal transcribed spacer region sequencing and cloning.

Fascioliasis causes high economic losses in livestock and underlies public health problems in rural areas, mainly of low-income countries. The increasing animal infection rates in Bangladesh were assessed, by focusing on host species, different parts of the country, and rDNA sequences. Fasciolid flukes were collected from buffaloes, cattle, goats and sheep from many localities to assess prevalences and intensities of infection. The nuclear rDNA internal transcribed spacer (ITS) region including ITS-1 and ITS-2 spacers was analyzed by direct sequencing and cloning, given the detection of intermediate phenotypic forms in Bangladesh. The 35.4% prevalence in goats and 55.5% in buffaloes are the highest recorded in these animals in Bangladesh. In cattle (29.3%) and sheep (26.8%) prevalences are also high for these species. These prevalences are very high when compared to lowlands at similar latitudes in neighboring India. The high prevalences and intensities appear in western Bangladesh where cross-border importation of animals from India occur. The combined haplotype CH3A of Fasciola gigantica widely found in all livestock species throughout Bangladesh fits its historical connections with the western Grand Trunk Road and the eastern Tea-Horse Road. The "pure" F. hepatica sequences only in clones from specimens showing heterozygotic positions indicate recent hybridization events with local "pure" F. gigantica, since concerted evolution did not yet have sufficient time to homogenize the rDNA operon. The detection of up to six different sequences coexisting in the cloned specimens evidences crossbreeding between hybrid parents, indicating repeated, superimposed and rapidly evolving hybridization events. The high proportion of hybrids highlights an increasing animal infection trend and human infection risk, and the need for control measures, mainly concerning goats in household farming management. ITS-1 and ITS-2 markers prove to be useful for detecting recent hybrid fasciolids. The introduction of a Fasciola species with imported livestock into a highly prevalent area of the other Fasciola species may lead to a high nucleotide variation in the species-differing positions in the extremely conserved fasciolid spacers. Results suggest that, in ancient times, frequent crossbreeding inside the same Fasciola species gave rise to the very peculiar characteristics of the present-day nuclear genome of both fasciolids.

Comparison of Serum Zinc in Children of Wilson Disease and Non-Wilsonian Volunteers in Bangladesh.

Wilson disease (WD) is an autosomal recessive disorder of copper metabolism with diverse clinical manifestations. Zinc (Zn) has been used for treatment of WD. Recent studies showed low serum zinc level in patients suffering from WD than the normal. This cross-sectional analytical study has been designed to compare the serum zinc level between paediatric patients suffering from WD but yet not started treatment and children who have normal ALT level. This study was carried out at the Department of Pediatric Gastroenterology and Nutrition, BSMMU, Dhaka, Bangladesh from July 2018 to June 2019. Total 51 children were included in this study. Among them 27 were diagnosed case of WD aged between three to eighteen years and 24 children of same ages who were suffering from other than liver disease having normal ALT were included as volunteers. The patients of WD were divided into four groups according to their presentation as acute hepatitis, chronic liver disease (CLD), acute liver failure & neuropsychiatric manifestation. Informed written consent was obtained from all patients and volunteers for participation in this study. Along with other physical findings and laboratory investigations 3 ml of venous blood were collected for estimation of serum zinc level. After estimation of serum zinc level results were analyzed statistically. The difference in serum zinc levels were compared between the groups. Serum zinc level was significantly lower in Wilson disease patients (43.8±19.7µg/dl; range: 13-83) compared to volunteers group (67.8±11.8µg/dl; range: 47-97) p<0.001. Among the diseased group, serum zinc level were significantly lower in 18 CLD (38.4±17.4µg/dl) and in 4 acute liver failure (33.1±3.7µg/dl) compared to 4 acute hepatitis (71.8±4.3µg/dl) (p=0.001) and (p<0.001) respectively. Mean serum zinc level was low in 4 Wilsonian acute liver failure (33.1±3.7µg/dl), which was significant compared to those (23) who presented as Wilson disease non acute liver failure (45.7±20.8µg/dl) (p=0.013). Serum zinc level was significantly lower in Wilson disease children compared to the volunteers. Zinc level was also found significantly low in Wilson disease presented as CLD and acute liver failure in comparison to Wilson disease presented as acute hepatitis.

Antibacterial effect of kitchen herbs against pathogenic multidrug-resistant E. coli isolates from calf diarrhoea.

Calf diarrhoea remains the biggest challenge both in the small and large farms. Infectious diarrhoea is associated with many pathogens, Escherichia coli being one, but majority are systematically treated with antibiotics. Since antimicrobial resistance (AMR) is a growing menace, the need to find alternative prophylactic solutions using popular kitchen herbs such as Trachyspermum ammi (carom seeds), Curcuma longa (turmeric) and cinnamon (Cinnamomum sp.) extracts is been investigated against virulent form of E. coli isolated from calf diarrhoea. The virulence factors identified in these isolates were ST (32.5%), LT (20%), eaeA (15%), stx1 (2.5%) and stx2 (5%) with the occurrence of the most common serogroups as O18 (15%) followed by O111 (12.5%). Highest resistance was seen with beta lactam + beta lactamase inhibitor (amoxicillin/clavulanic acid) followed by beta lactams (ampicillin, cefuroxime and cefepime). The zone of inhibition due to cinnamon (methanol) and carom seed (ethanol) extracts (500 to 250 µg/mL concentration) on E. coli bacteria was >19 mm, respectively. Turmeric, cinnamon and carom had the potency of inhibiting the pathogenic E. coli which maybe suggestive of its use in calf diets as prophylaxis against diarrhoea.

Study of Body Mass Index, Blood Pressure, Serum Sodium in Newly Diagnosed Hypertensive Patients.

Hypertension 'The sustained elevation of systemic arterial pressure' is a significant risk factor for heart disease, stroke and other cardiovascular diseases and an estimated 970 million people worldwide suffer from the disease resulting in significant morbidity, mortality and financial burden globally. It is the leading modifiable risk factor for morbidity and mortality worldwide. Worldwide an estimated 1.28 billion adults aged 30-79 years have hypertension, the majority (two-thirds) living with low and middle income countries. One of the global targets for non-communicable diseases is to reduce the prevalence of hypertension by 33% between 2010 and 2030.Sodium plays an important role in blood pressure regulation with a reduced sodium intake being associated with a reduction in systolic and diastolic blood pressure. This study was done to evaluate the differences in body mass index (BMI) and serum sodium in hypertensive and normotensive subjects. This analytical type of cross sectional study was carried out in the Department of Physiology, Mymensingh Medical College, Mymensingh between the periods from January 2022 to December 2022. A total number of 140 male subjects, age ranged from (30-59) years were included in this study. Among them, seventy (70) hypertensive subjects were taken as study group (Group II) and seventy (70) age matched normotensive subjects were taken as control group (Group I). The results were calculated and analyzed by using Statistical Package for Social Science (SPSS) version 26.0. Anthropometric measurements like height and weight taken in meter and kilogram respectively. Systolic and diastolic blood pressure was performed by aneroid sphygmomanometer (ALPK2, Japan), laboratory analysis of serum sodium by colorimetric method. In this study we found that body mass index in control group 23.59±1.29 kg/m² and study group 26.81±2.31kg/m²; blood pressure (systolic pressure in control group 113.21±6.76 mm Hg and in study group 149.14±5.03 mm Hg, diastolic pressure in control group 75.57±4.55 mm Hg and in study group 100.21±5.28 mm Hg) and serum sodium in control group 138.84±2.12 & in study group was 147.94±1.41 which were significant in study group in comparison with control group. In study group parameters were significantly increased in comparison to control male group. Therefore, by this study we recommended that routine estimation of these parameters is important for prevention of complication related to hypertension for leading a healthy life.

LDDNet: A Deep Learning Framework for the Diagnosis of Infectious Lung Diseases.

This paper proposes a new deep learning (DL) framework for the analysis of lung diseases, including COVID-19 and pneumonia, from chest CT scans and X-ray (CXR) images. This framework is termed optimized DenseNet201 for lung diseases (LDDNet). The proposed LDDNet was developed using additional layers of 2D global average pooling, dense and dropout layers, and batch normalization to the base DenseNet201 model. There are 1024 Relu-activated dense layers and 256 dense layers using the sigmoid activation method. The hyper-parameters of the model, including the learning rate, batch size, epochs, and dropout rate, were tuned for the model. Next, three datasets of lung diseases were formed from separate open-access sources. One was a CT scan dataset containing 1043 images. Two X-ray datasets comprising images of COVID-19-affected lungs, pneumonia-affected lungs, and healthy lungs exist, with one being an imbalanced dataset with 5935 images and the other being a balanced dataset with 5002 images. The performance of each model was analyzed using the Adam, Nadam, and SGD optimizers. The best results have been obtained for both the CT scan and CXR datasets using the Nadam optimizer. For the CT scan images, LDDNet showed a COVID-19-positive classification accuracy of 99.36%, a 100% precision recall of 98%, and an F1 score of 99%. For the X-ray dataset of 5935 images, LDDNet provides a 99.55% accuracy, 73% recall, 100% precision, and 85% F1 score using the Nadam optimizer in detecting COVID-19-affected patients. For the balanced X-ray dataset, LDDNet provides a 97.07% classification accuracy. For a given set of parameters, the performance results of LDDNet are better than the existing algorithms of ResNet152V2 and XceptionNet.

Machine learning for DCO-OFDM based LiFi.

Light fidelity (LiFi) uses different forms of orthogonal frequency division multiplexing (OFDM), including DC biased optical OFDM (DCO-OFDM). In DCO-OFDM, the use of a large DC bias causes optical power inefficiency, while a small bias leads to higher clipping noise. Hence, finding an appropriate DC bias level for DCO-OFDM is important. This paper applies machine learning (ML) algorithms to find optimum DC-bias value for DCO-OFDM based LiFi systems. For this, a dataset is generated for DCO-OFDM using MATLAB tool. Next, ML algorithms are applied using Python programming language. ML is used to find the important attributes of DCO-OFDM that influence the optimum DC bias. It is shown here that the optimum DC bias is a function of several factors including, the minimum, the standard deviation, and the maximum value of the bipolar OFDM signal, and the constellation size. Next, linear and polynomial regression algorithms are successfully applied to predict the optimum DC bias value. Results show that polynomial regression of order 2 can predict the optimum DC bias value with a coefficient of determination of 96.77% which confirms the effectiveness of the prediction.

CO-IRv2: Optimized InceptionResNetV2 for COVID-19 detection from chest CT images.

This paper focuses on the application of deep learning (DL) in the diagnosis of coronavirus disease (COVID-19). The novelty of this work is in the introduction of optimized InceptionResNetV2 for COVID-19 (CO-IRv2) method. A part of the CO-IRv2 scheme is derived from the concepts of InceptionNet and ResNet with hyperparameter tuning, while the remaining part is a new architecture consisting of a global average pooling layer, batch normalization, dense layers, and dropout layers. The proposed CO-IRv2 is applied to a new dataset of 2481 computed tomography (CT) images formed by collecting two independent datasets. Data resizing and normalization are performed, and the evaluation is run up to 25 epochs. Various performance metrics, including precision, recall, accuracy, F1-score, area under the receiver operating characteristics (AUC) curve are used as performance metrics. The effectiveness of three optimizers known as Adam, Nadam and RMSProp are evaluated in classifying suspected COVID-19 patients and normal people. Results show that for CO-IRv2 and for CT images, the obtained accuracies of Adam, Nadam and RMSProp optimizers are 94.97%, 96.18% and 96.18%, respectively. Furthermore, it is shown here that for the case of CT images, CO-IRv2 with Nadam optimizer has better performance than existing DL algorithms in the diagnosis of COVID-19 patients. Finally, CO-IRv2 is applied to an X-ray dataset of 1662 images resulting in a classification accuracy of 99.40%.

Demographic and Clinical Profile of COVID-19 ICU Patients of Bangabandhu Sheikh Mujib Medical University, Bangladesh.

Corona viruses are a group of RNA viruses that cause infection in humans and animals. In human Corona viruses cause respiratory tract infections ranging from mild to critical illness. Corona virus disease 2019 (COVID-19) is caused by strain of severe acute respiratory syndrome corona virus 2 (SARS-CoV-2). The disease was first identified in Wuhan city, of China, in December 2019 and since spread all around the world. In Bangladesh first case has been declared by Institute of Epidemiology, Disease Control and Research (IEDCR) in 8th March, 2020 and first death on 18th march in an ICU and by 13th December total 489,178 cases and 7,020 deaths occurred in this country. The main objective of this study was to determine the Demographic and Clinical Profile of COVID-19 ICU patients in Bangladesh. This retrospective descriptive study on clinical profile along with short term treatment outcomes of COVID-19 patients conducted from COVID-19 dedicated Intensive care unit of Bangabandhu Sheikh Mujib Medical University (BSMMU), Bangladesh during July 2020 to November 2020. Total 300 ICU patients was included in this study. The age range of the patients was from 15 to 91 years. The highest percentage of patients about 49.00%, which was 147 patients were belonged to 61-75 years age group. The patients mean age was 62.80 years. Regarding gender distribution, among those 300 patients; 77.00% which is 231 were male and only 23.00% which is 69 were female. Patients admitted with symptoms like, respiratory distress/shortness of breath (100.00%), persistent worsening cough (60.00%), fatigue (55.00%) and fever (40.00%). Patients were also present with sore throat (35.00%), rhinorrhea (30.00%), altered mental status (20.00%), diarrhoea (10.00%) and chest pain (5.00%). Regarding co-morbidities, around half of the patients were suffering from Diabetes (60.22%) and Hypertension (53.44%). Significant amount of patients were also suffering from chronic obstructive pulmonary disease (27.00%) and bronchial asthma (16.78%). Ischemic heart disease was (10.33%), chronic kidney disease (10.89%), hypothyroidism (9.78%) and multiple co-morbidities (15.12%) at the time of admission. Mortality rate in this case were 71.00% and most of the death cases were in between 61 to 75 years of age group (40.00%). After improvement 27.00% patients were transferred to cabin for further management. We could discharge to home directly only 2.00% of patients.

Differences in the socio-demographic determinants of undernutrition in children aged <5 years in urban and rural areas of Bangladesh measured by the Composite Index of Anthropometric Failure.

OBJECTIVES: This study investigates the prevalence and determinants of undernutrition among children <5 years living in Bangladesh using the Composite Index of Anthropometric Failure (CIAF) and highlights the differences between urban and rural areas. STUDY DESIGN: Data are drawn from three cross-sectional Bangladesh Demographic Health Surveys conducted from 2007 to 2014. METHODS: A Chi-square test was used to assess the prevalence of <5 years child undernutrition. Logistic regression analysis was performed to identify various sociodemographic risk factors. RESULTS: The prevalence of undernutrition based on the CIAF was 52% among children <5 years in Bangladesh. The prevalence of undernutrition in children living in urban areas and rural areas were found to be 45% and 54%, respectively. As per the CIAF, undernutrition was highly prevalent among children in the older age group, children of uneducated and currently working mothers, those of underweight mothers, children of fourth and above in the birth order, children of fathers who were manual labourers, children of households who had no access to television and those in the poorest households whether in urban or rural areas. Children in the older age group, children of uneducated mothers, those with underweight mothers and those from the poorest households provided common key risk factors for undernutrition in both urban and rural areas. Children of fourth and above birth order and not watching television at all were additional risk factors of child undernutrition in rural areas. CONCLUSION: Half of the children in rural areas and two-fifths of them in urban areas are suffering undernutrition in Bangladesh, and several sociodemographic factors heighten the risks. Also, birth order and watching television were identified as the differential risk factors. This study therefore concludes that evidence-based interventions are needed to reduce the burden of undernutrition in children in the country.

Diagnosis of COVID-19 Using Machine Learning and Deep Learning: A Review.

BACKGROUND: This paper provides a systematic review of the application of Artificial Intelligence (AI) in the form of Machine Learning (ML) and Deep Learning (DL) techniques in fighting against the effects of novel coronavirus disease (COVID-19). OBJECTIVE & METHODS: The objective is to perform a scoping review on AI for COVID-19 using preferred reporting items of systematic reviews and meta-analysis (PRISMA) guidelines. A literature search was performed for relevant studies published from 1 January 2020 till 27 March 2021. Out of 4050 research papers available in reputed publishers, a full-text review of 440 articles was done based on the keywords of AI, COVID-19, ML, forecasting, DL, X-ray, and Computed Tomography (CT). Finally, 52 articles were included in the result synthesis of this paper. As part of the review, different ML regression methods were reviewed first in predicting the number of confirmed and death cases. Secondly, a comprehensive survey was carried out on the use of ML in classifying COVID-19 patients. Thirdly, different datasets on medical imaging were compared in terms of the number of images, number of positive samples and number of classes in the datasets. The different stages of the diagnosis, including preprocessing, segmentation and feature extraction were also reviewed. Fourthly, the performance results of different research papers were compared to evaluate the effectiveness of DL methods on different datasets. RESULTS: Results show that residual neural network (ResNet-18) and densely connected convolutional network (DenseNet 169) exhibit excellent classification accuracy for X-ray images, while DenseNet-201 has the maximum accuracy in classifying CT scan images. This indicates that ML and DL are useful tools in assisting researchers and medical professionals in predicting, screening and detecting COVID-19. CONCLUSION: Finally, this review highlights the existing challenges, including regulations, noisy data, data privacy, and the lack of reliable large datasets, then provides future research directions in applying AI in managing COVID-19.

Microplastic particles in the aquatic environment: A systematic review.

Microplastics (MPs) pollution has become one of the most severe environmental concerns today. MPs persist in the environment and cause adverse effects in organisms. This review aims to present a state-of-the-art overview of MPs in the aquatic environment. Personal care products, synthetic clothing, air-blasting facilities and drilling fluids from gas-oil industries, raw plastic powders from plastic manufacturing industries, waste plastic products and wastewater treatment plants act as the major sources of MPs. For MPs analysis, pyrolysis-gas chromatography-mass spectrometry (Py-GC-MS), Py-MS methods, Raman spectroscopy, and FT-IR spectroscopy are regarded as the most promising methods for MPs identification and quantification. Due to the large surface area to volume ratio, crystallinity, hydrophobicity and functional groups, MPs can interact with various contaminants such as heavy metals, antibiotics and persistent organic contaminants. Among different physical and biological treatment technologies, the MPs removal performance decreases as membrane bioreactor (> 99%) > activated sludge process (~98%) > rapid sand filtration (~97.1%) > dissolved air floatation (~95%) > electrocoagulation (> 90%) > constructed wetlands (88%). Chemical treatment methods such as coagulation, magnetic separations, Fenton, photo-Fenton and photocatalytic degradation also show moderate to high efficiency of MP removal. Hybrid treatment technologies show the highest removal efficacies of MPs. Finally, future research directions for MPs are elaborated.

Data analytics for novel coronavirus disease.

This paper describes different aspects of novel coronavirus disease (COVID-19), presents visualization of the spread of the infection, and discusses the potential applications of data analytics on this viral infection. Firstly, a literature survey is done on COVID-19 highlighting a number of factors including its origin, its similarity with previous coronaviruses, its transmission capacity, its symptoms, etc. Secondly, data analytics is applied on a dataset of Johns Hopkins University to find out the spread of the viral infection. It is shown here that although the disease started in China in December 2019, the highest number of confirmed cases up to June 04, 2020 is in the USA. Thirdly, the worldwide increase in the number of confirmed cases over time is modelled here using a polynomial regression algorithm with degree 2. Fourthly, classification algorithms are applied on a dataset of 5644 samples provided by Hospital Israelita Albert Einstein of Brazil in order to diagnose COVID-19. It is shown here that multilayer perceptron (MLP), XGBoost and logistic regression can classify COVID-19 patients at an accuracy above 91%. Finally, a discussion is presented on the potential applications of data analytics in several important factors of COVID-19.

Hybrid deep learning for detecting lung diseases from X-ray images.

Lung disease is common throughout the world. These include chronic obstructive pulmonary disease, pneumonia, asthma, tuberculosis, fibrosis, etc. Timely diagnosis of lung disease is essential. Many image processing and machine learning models have been developed for this purpose. Different forms of existing deep learning techniques including convolutional neural network (CNN), vanilla neural network, visual geometry group based neural network (VGG), and capsule network are applied for lung disease prediction. The basic CNN has poor performance for rotated, tilted, or other abnormal image orientation. Therefore, we propose a new hybrid deep learning framework by combining VGG, data augmentation and spatial transformer network (STN) with CNN. This new hybrid method is termed here as VGG Data STN with CNN (VDSNet). As implementation tools, Jupyter Notebook, Tensorflow, and Keras are used. The new model is applied to NIH chest X-ray image dataset collected from Kaggle repository. Full and sample versions of the dataset are considered. For both full and sample datasets, VDSNet outperforms existing methods in terms of a number of metrics including precision, recall, F0.5 score and validation accuracy. For the case of full dataset, VDSNet exhibits a validation accuracy of 73%, while vanilla gray, vanilla RGB, hybrid CNN and VGG, and modified capsule network have accuracy values of 67.8%, 69%, 69.5% and 63.8%, respectively. When sample dataset rather than full dataset is used, VDSNet requires much lower training time at the expense of a slightly lower validation accuracy. Hence, the proposed VDSNet framework will simplify the detection of lung disease for experts as well as for doctors.

Data-driven diagnosis of spinal abnormalities using feature selection and machine learning algorithms.

This paper focuses on the application of machine learning algorithms for predicting spinal abnormalities. As a data preprocessing step, univariate feature selection as a filter based feature selection, and principal component analysis (PCA) as a feature extraction algorithm are considered. A number of machine learning approaches namely support vector machine (SVM), logistic regression (LR), bagging ensemble methods are considered for the diagnosis of spinal abnormality. The SVM, LR, bagging SVM and bagging LR models are applied on a dataset of 310 samples publicly available in Kaggle repository. The performance of classification of abnormal and normal spinal patients is evaluated in terms of a number of factors including training and testing accuracy, recall, and miss rate. The classifier models are also evaluated by optimizing the kernel parameters, and by using the results of receiver operating characteristic (ROC) and precision-recall curves. Results indicate that when 78% data are used for training, the observed training accuracies for SVM, LR, bagging SVM and bagging LR are 86.30%, 85.47%, 86.72% and 85.06%, respectively. On the other hand, the accuracies for the test dataset for SVM, LR, bagging SVM and bagging LR are the same being 86.96%. However, bagging SVM is the most attractive as it has a higher recall value and a lower miss rate compared to others. Hence, bagging SVM is suitable for the classification of spinal patients when applied on the most five important features of spinal samples.

Molecular Identification of Vibrio alginolyticus Causing Vibriosis in Shrimp and Its Herbal Remedy.

Penaeus monodon is highly susceptible to vibriosis disease. Aims of the study were to identify the pathogen causing vibriosis in P. monodon through molecular techniques and develop a biocontrol method of the disease by application of herbal extracts. Shrimp samples were collected aseptically from the infected farm and the bacteria were isolated from the infected region of those samples. Based on phenotypic identification, several isolates were identified as Vibrio sp. 16S rRNA gene sequences of the selected isolates exhibited 100% homology with V. alginolyticus strain ATCC 17749. An in vivo infection challenge test was performed by immersion method with V. alginolyticus where these isolates caused high mortality in juvenile shrimp with prominent symptoms of hepatopancreatic necrosis. Antibiogram profile of the isolates was determined against eleven commercial antibiotic discs whereas the isolates were found resistant to multiple antibiotics. A total of twenty-one herbal extracts were screened where Emblica officinalis, Allium sativum, and Syzygium aromaticum strongly inhibited the growth of V. alginolyticus in in vitro conditions. In in vivo conditions, the ethyl acetate extracts of E. officinalis and A. sativum successfully controlled the vibriosis disease in shrimp at a dose of 10 mg/g feed. This is the first report on molecular identification and biocontrol of V. alginolyticus in shrimp in Bangladesh.Penaeus monodon is highly susceptible to vibriosis disease. Aims of the study were to identify the pathogen causing vibriosis in P. monodon through molecular techniques and develop a biocontrol method of the disease by application of herbal extracts. Shrimp samples were collected aseptically from the infected farm and the bacteria were isolated from the infected region of those samples. Based on phenotypic identification, several isolates were identified as Vibrio sp. 16S rRNA gene sequences of the selected isolates exhibited 100% homology with V. alginolyticus strain ATCC 17749. An in vivo infection challenge test was performed by immersion method with V. alginolyticus where these isolates caused high mortality in juvenile shrimp with prominent symptoms of hepatopancreatic necrosis. Antibiogram profile of the isolates was determined against eleven commercial antibiotic discs whereas the isolates were found resistant to multiple antibiotics. A total of twenty-one herbal extracts were screened where Emblica officinalis, Allium sativum, and Syzygium aromaticum strongly inhibited the growth of V. alginolyticus in in vitro conditions. In in vivo conditions, the ethyl acetate extracts of E. officinalis and A. sativum successfully controlled the vibriosis disease in shrimp at a dose of 10 mg/g feed. This is the first report on molecular identification and biocontrol of V. alginolyticus in shrimp in Bangladesh.

Renal Involvement as a Presenting Feature of Multiple Myeloma.

Renal involvement may be the presenting feature in a vast majority of patients with multiple myeloma and is one of the key for clinical manifestations of symptomatic multiple myeloma. The purpose of the study was to find out the pattern of renal involvement at the time of presentation of multiple myeloma and to explore its association with clinical, laboratory and pathologic features of these cases. This cross sectional study was conducted in the Department of Nephrology at Bangabandhu Sheikh Mujib Medical University (BSMMU), Dhaka, Bangladesh from February 2016 to September 2017. Forty seven (47) patients of newly diagnosed multiple myeloma having renal involvement were included in the study. Multiple myeloma was diagnosed as per criteria proposed by the International Myeloma Working Group, 2003. Renal involvement was considered to be present when any one of proteinuria, microscopic haematuria, renal impairment or urinary tract infection (UTI) was found in the patient. Renal biopsy was done in suitable patients under ultrasound guidance after taking informed written consent. The pattern of renal involvement was detected and status of renal function was assessed and its clinical, laboratory and pathologic associations were analyzed. Data were managed by using computer based software, the Statistical Package for Social Sciences (SPSS) version 23 (IBM Corp.). Median age at presentation was 59 years with the range of 37-76 years. Female (53.2%) was slightly predominant than male (46.8%) and male to female ratio was 1:1.14. Renal impairment, microscopic haematuria, proteinuria, nephrotic range proteinuria, urinary Bence Jones protein and UTI were found in 70%, 19%, 79%, 25%, 19% and 17% of patients respectively. Median serum creatinine and proteinuria were 256µmol/l and 1.24gm/day. Hypercalcaemia and Bence Jones proteinuria were detected in 36% and 27% of patients respectively with renal impairment which were statistically significant. The precipitating factors for renal impairment were NSAIDs use (67%), hyperuricaemia (49%), hypercalcaemia (36%), dehydration (27%), UTI (18%) and no identifiable factor (3%). Dialysis was required in 15% new myeloma patient. Renal biopsy and histopathological examination revealed myeloma cast nephropathy (30%), amyloidosis (30%), glomerulosclerosis (chronic kidney disease) (20%), monoclonal immunoglobulin deposition disease (MIDD) (10%) and interstitial nephritis with fibrosis (10%). Renal involvement was a common and severe complication of multiple myeloma. Renal impairment was strongly associated with hypercalcaemia, NSAIDs use, hyperuricaemia, Bence Jones proteinuria etc.