Singular-Value-Decomposition-Based Matrix Surgery.

This paper is motivated by the need to stabilise the impact of deep learning (DL) training for medical image analysis on the conditioning of convolution filters in relation to model overfitting and robustness. We present a simple strategy to reduce square matrix condition numbers and investigate its effect on the spatial distributions of point clouds of well- and ill-conditioned matrices. For a square matrix, the SVD surgery strategy works by: (1) computing its singular value decomposition (SVD), (2) changing a few of the smaller singular values relative to the largest one, and (3) reconstructing the matrix by reverse SVD. Applying SVD surgery on CNN convolution filters during training acts as spectral regularisation of the DL model without requiring the learning of extra parameters. The fact that the further away a matrix is from the non-invertible matrices, the higher its condition number is suggests that the spatial distributions of square matrices and those of their inverses are correlated to their condition number distributions. We shall examine this assertion empirically by showing that applying various versions of SVD surgery on point clouds of matrices leads to bringing their persistent diagrams (PDs) closer to the matrices of the point clouds of their inverses.

An Evaluation of the Effectiveness of Image-based Texture Features Extracted from Static B-mode Ultrasound Images in Distinguishing between Benign and Malignant Ovarian Masses.

Significant successes in machine learning approaches to image analysis for various applications have energized strong interest in automated diagnostic support systems for medical images. The evolving in-depth understanding of the way carcinogenesis changes the texture of cellular networks of a mass/tumor has been informing such diagnostics systems with use of more suitable image texture features and their extraction methods. Several texture features have been recently applied in discriminating malignant and benign ovarian masses by analysing B-mode images from ultrasound scan of the ovary with different levels of performance. However, comparative performance evaluation of these reported features using common sets of clinically approved images is lacking. This paper presents an empirical evaluation of seven commonly used texture features (histograms, moments of histogram, local binary patterns [256-bin and 59-bin], histograms of oriented gradients, fractal dimensions, and Gabor filter), using a collection of 242 ultrasound scan images of ovarian masses of various pathological characteristics. The evaluation examines not only the effectiveness of classification schemes based on the individual texture features but also the effectiveness of various combinations of these schemes using the simple majority-rule decision level fusion. Trained support vector machine classifiers on the individual texture features without any specific pre-processing, achieve levels of accuracy between 75% and 85% where the seven moments and the 256-bin LBP are at the lower end while the Gabor filter is at the upper end. Combining the classification results of the top k (k = 3, 5, 7) best performing features further improve the overall accuracy to a level between 86% and 90%. These evaluation results demonstrate that each of the investigated image-based texture features provides informative support in distinguishing benign or malignant ovarian masses.

Bacteriophage biocontrol in wastewater treatment.

Waterborne bacterial pathogens in wastewater remains an important public health concern, not only because of the environmental damage, morbidity and mortality that they cause, but also due to the high cost of disinfecting wastewater by using physical and chemical methods in treatment plants. Bacteriophages are proposed as bacterial pathogen indicators and as an alternative biological method for wastewater treatment. Phage biocontrol in large scale treatment requires adaptive and aggressive phages that are able to overcome the environmental forces that interfere with phage-host interactions while targeting unwanted bacterial pathogens and preventing biofilms and foaming. This review will shed light on aspects of using bacteriophage programming technology in wastewater plants to rapidly target and reduce undesirable bacteria without harming the useful bacteria needed for biodegradation.

The potential of bacteriophage cocktail in eliminating Methicillin-resistant Staphylococcus aureus biofilms in terms of different extracellular matrices expressed by PIA, ciaA-D and FnBPA genes.

BACKGROUND: This study assessed novel approach of using highly lytic phages against methicillin-susceptible Staphylococcus aureus (MSSA) and methicillin-resistant Staphylococcus aureus (MRSA) biofilms with and without biofilm extracellular matrix- disrupting chemical. METHOD: The resultant phage-based control was assessed in relation to the type of biofilm extracellular matrix namely, polysaccharide intercellular adhesion (PIA) or proteinacious fibronectin-binding protein A (FnBPA). The biofilms were formed in vitro by 24 h incubation of bacteria in 96 wells microtiter plates at room temperature. The formed biofilms were assessed by tissue culture plate (TCP). Moreover, the nature of the biofilm was assessed by scanning electron microscopy (SEM) and PCR assay for detecting PIA genes, ciaA-D and FnBPA genes. RESULTS: this study showed that applied phages with 0.08 % benezenthonium chloride, for PIA biofilms, and 0.06 % ethanol, for proteinacious FnBPA biofilms, exerted 100 % eradication for MSSA biofilms and about 78 % of MRSA biofilms. The phage-based control of biofilms with chemical adjuvant showed significantly higher efficiency than that without adjuvant (P < 0.05). Moreover, FnBPA biofilms were more common in MRSA than in MSSA while PIA biofilms were more common in MSSA than in MRSA. And the most resistant type of biofilms to phage-based control was FnBPA in MRSA where 50 % of biofilms were reduced but not eradicated completely. CONCLUSIONS: It is concluded that PIA-disturbing agent and protein denaturing alcohol can increase the efficiency of attacking phages in accessing host cell walls and lysing them which in turn lead to much more efficient MRSA and MSSA biofilm treatment and prevention.

A MATLAB tool for pathway enrichment using a topology-based pathway regulation score.

BACKGROUND: Handling the vast amount of gene expression data generated by genome-wide transcriptional profiling techniques is a challenging task, demanding an informed combination of pre-processing, filtering and analysis methods if meaningful biological conclusions are to be drawn. For example, a range of traditional statistical and computational pathway analysis approaches have been used to identify over-represented processes in microarray data derived from various disease states. However, most of these approaches tend not to exploit the full spectrum of gene expression data, or the various relationships and dependencies. Previously, we described a pathway enrichment analysis tool created in MATLAB that yields a Pathway Regulation Score (PRS) by considering signalling pathway topology, and the overrepresentation and magnitude of differentially-expressed genes (J Comput Biol 19:563-573, 2012). Herein, we extended this approach to include metabolic pathways, and described the use of a graphical user interface (GUI). RESULTS: Using input from a variety of microarray platforms and species, users are able to calculate PRS scores, along with a corresponding z-score for comparison. Further pathway significance assessment may be performed to increase confidence in the pathways obtained, and users can view Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway diagrams marked-up to highlight impacted genes. CONCLUSIONS: The PRS tool provides a filter in the isolation of biologically-relevant insights from complex transcriptomic data.

Novel approach of using a cocktail of designed bacteriophages against gut pathogenic E. coli for bacterial load biocontrol.

BACKGROUND: This study was conducted to explore new approaches of animal biocontrol via biological control feed. METHOD: White rats were subjected to 140 highly lytic designed phages specific against E. coli. Phages were fed via drinking water, oral injection, and vegetable capsules. Phage feeding was applied by 24 h feeding with 11 d monitoring and 20 d phage feeding and monitoring. Group of rats received external pathogenic E. coli and another group did not, namely groups A and B. RESULTS: Phage feeding for 20 d via vegetable capsules yielded the highest reduction of fecal E. coli, 3.02 and 4.62 log, in rats group A and B respectively. Second best, feeding for 20 d via drinking water with alkali yielded 2.78 and 4.08 log in rats groups A and B respectively. The peak reduction in E. coli output was 5-10 d after phage feeding. Phage control declined after 10th day of feeding. CONCLUSIONS: The use of cocktail of designed phages succeeded in suppressing flora or external E. coli. The phage feed biocontrol is efficient in controlling E. coli at the pre-harvest period, precisely at the 6th-8th day of phage feeding when the lowest E. coli output found.

Natural solution to antibiotic resistance: bacteriophages 'The Living Drugs'.

Antibiotics have been a panacea in animal husbandry as well as in human therapy for decades. The huge amount of antibiotics used to induce the growth and protect the health of farm animals has lead to the evolution of bacteria that are resistant to the drug's effects. Today, many researchers are working with bacteriophages (phages) as an alternative to antibiotics in the control of pathogens for human therapy as well as prevention, biocontrol, and therapy in animal agriculture. Phage therapy and biocontrol have yet to fulfill their promise or potential, largely due to several key obstacles to their performance. Several suggestions are shared in order to point a direction for overcoming common obstacles in applied phage technology. The key to successful use of phages in modern scientific, farm, food processing and clinical applications is to understand the common obstacles as well as best practices and to develop answers that work in harmony with nature.

A novel method to assess collagen architecture in skin.

BACKGROUND: Texture within biological specimens may reveal critical insights, while being very difficult to quantify. This is a particular problem in histological analysis. For example, cross-polar images of picrosirius stained skin reveal exquisite structure, allowing changes in the basketweave conformation of healthy collagen to be assessed. Existing techniques measure gross pathological changes, such as fibrosis, but are not sufficiently sensitive to detect more subtle and progressive pathological changes in the dermis, such as those seen in ageing. Moreover, screening methods for cutaneous therapeutics require accurate, unsupervised and high-throughput image analysis techniques. RESULTS: By analyzing spectra of images post Gabor filtering and Fast Fourier Transform, we were able to measure subtle changes in collagen fibre orientation intractable to existing techniques. We detected the progressive loss of collagen basketweave structure in a series of chronologically aged skin samples, as well as in skin derived from a model of type 2 diabetes mellitus. CONCLUSIONS: We describe a novel bioimaging approach with implications for the evaluation of pathology in a broader range of biological situations.

Impact of external forces on cyanophage-host interactions in aquatic ecosystems.

Cyanobacterial (algal) blooms have by convention been attributed to the excessive level of nutrients from pollution and runoff, which promotes the rapid growth and multiplication of cyanobacteria or algae. The cyanophage (virus) is the natural predator of cyanobacteria (the host). The aim of this review is to unveil certain pressures that disrupt cyanophage-host interactions and the formation of cyanobacterial blooms. This review focuses principally on the impact of greenhouse gases, ozone depletion, solar ultraviolet radiation (SUR) and the role of recently discovered virophages, which coexist with and in turn are the natural predator of phages. The key findings are that the increase in SUR, the mutation of cyanophages and cyanobacteria, along with changing nutrient levels, have combined with virophages to impede cyanophage-host interactions and the resultant viral infection and killing of the cyanobacterial cell, which is a necessary step in controlling cyanobacterial blooms. Consider this a 'call to action' for researchers interested in corrective action aimed at evolving aquatic ecosystems.

A generic deep learning framework to classify thyroid and breast lesions in ultrasound images.

Breast and thyroid cancers are the two common cancers to affect women worldwide. Ultrasonography (US) is a commonly used non-invasive imaging modality to detect breast and thyroid cancers, but its clinical diagnostic accuracy for these cancers is controversial. Both thyroid and breast cancers share some similar high frequency ultrasound characteristics such as taller-than-wide shape ratio, hypo-echogenicity, and ill-defined margins. This study aims to develop an automatic scheme for classifying thyroid and breast lesions in ultrasound images using deep convolutional neural networks (DCNN). In particular, we propose a generic DCNN architecture with transfer learning and the same architectural parameter settings to train models for thyroid and breast cancers (TNet and BNet) respectively, and test the viability of such a generic approach with ultrasound images collected from clinical practices. In addition, the potentials of the thyroid model in learning the common features and its performance of classifying both breast and thyroid lesions are investigated. A retrospective dataset of 719 thyroid and 672 breast images captured from US machines of different makes between October 2016 and December 2018 is used in this study. Test results show that both TNet and BNet built on the same DCNN architecture have achieved good classification results (86.5% average accuracy for TNet and 89% for BNet). Furthermore, we used TNet to classify breast lesions and the model achieves sensitivity of 86.6% and specificity of 87.1%, indicating its capability in learning features commonly shared by thyroid and breast lesions. We further tested the diagnostic performance of the TNet model against that of three radiologists. The area under curve (AUC) for thyroid nodule classification is 0.861 (95% CI: 0.792-0.929) for the TNet model and 0.757-0.854 (95% CI: 0.658-0.934) for the three radiologists. The AUC for breast cancer classification is 0.875 (95% CI: 0.804-0.947) for the TNet model and 0.698-0.777 (95% CI: 0.593-0.872) for the radiologists, indicating the model's potential in classifying both breast and thyroid cancers with a higher level of accuracy than that of radiologists.

In vitro Evaluation of the Antiviral Activity of an Extract of Date Palm (Phoenix dactylifera L.) Pits on a Pseudomonas Phage.

A crude acetone extract of the pit of date palm (Phoenix dactylifera L.) was prepared and its antiviral activity evaluated against lytic Pseudomonas phage ATCC 14209-B1, using Pseudomonas aeruginosa ATCC 25668 as the host cell. The antiviral activity of date pits was found to be mediated by binding to the phage, with minimum inhibitory concentration (MIC) of <10 µg ml(-1). The decimal reduction time (D-values), the concentration exponent (η) and the phage inactivation kinetics were determined. The date pit extracts show a strong ability to inhibit the infectivity of Pseudomonas phage ATCC 14209-B1 and completely prevented bacterial lysis, which it is hoped will promote research into its potential as a novel antiviral agent against pathogenic human viruses.

Evaluation of machine learning methods with Fourier Transform features for classifying ovarian tumors based on ultrasound images.

INTRODUCTION: Ovarian tumors are the most common diagnostic challenge for gynecologists and ultrasound examination has become the main technique for assessment of ovarian pathology and for preoperative distinction between malignant and benign ovarian tumors. However, ultrasonography is highly examiner-dependent and there may be an important variability between two different specialists when examining the same case. The objective of this work is the evaluation of different well-known Machine Learning (ML) systems to perform the automatic categorization of ovarian tumors from ultrasound images. METHODS: We have used a real patient database whose input features have been extracted from 348 images, from the IOTA tumor images database, holding together with the class labels of the images. For each patient case and ultrasound image, its input features have been previously extracted using Fourier descriptors computed on the Region Of Interest (ROI). Then, four ML techniques are considered for performing the classification stage: K-Nearest Neighbors (KNN), Linear Discriminant (LD), Support Vector Machine (SVM) and Extreme Learning Machine (ELM). RESULTS: According to our obtained results, the KNN classifier provides inaccurate predictions (less than 60% of accuracy) independently of the size of the local approximation, whereas the classifiers based on LD, SVM and ELM are robust in this biomedical classification (more than 85% of accuracy). CONCLUSIONS: ML methods can be efficiently used for developing the classification stage in computer-aided diagnosis systems of ovarian tumor from ultrasound images. These approaches are able to provide automatic classification with a high rate of accuracy. Future work should aim at enhancing the classifier design using ensemble techniques. Another ongoing work is to exploit different kind of features extracted from ultrasound images.

Application of a novel immunomagnetic separation-bacteriophage assay for the detection of Salmonella enteritidis and Escherichia coli O157:H7 in food.

Salmonella infection is the second most prevalent cause of foodborne illness in most developing countries. Meat, poultry, and dairy products are frequently implicated in outbreaks. The objective of this study was to apply a novel immunomagnetic separation (IMS)-bacteriophage assay to the detection of Salmonella enteritidis in artificially inoculated skimmed milk powder, chicken rinses, and ground beef. In all food types tested, the IMS-bacteriophage assay was able to detect an average of 3 CFU of S. enteritidis in 25 g or ml of food sample. Total assay time including pre-enrichment is about 20 h. The results indicate that the IMS-bacteriophage assay is a rapid and sensitive means of detecting S. enteritidis in these foods. The assay was successfully adapted to the detection of Escherichia coli O157:H7 and was able to detect E. coli in ground beef at the lowest inoculation level tested, 2 CFU/g. The assay was also adapted to the simultaneous detection of S. enteritidis and E. coli. The results indicate that the IMS-bacteriophage assay shows promise for the simultaneous detection of these pathogens, but further development work would be necessary to improve sensitivity and produce reliable results at low inoculation levels.

A novel automated image analysis method for accurate adipocyte quantification.

Increased adipocyte size and number are associated with many of the adverse effects observed in metabolic disease states. While methods to quantify such changes in the adipocyte are of scientific and clinical interest, manual methods to determine adipocyte size are both laborious and intractable to large scale investigations. Moreover, existing computational methods are not fully automated. We, therefore, developed a novel automatic method to provide accurate measurements of the cross-sectional area of adipocytes in histological sections, allowing rapid high-throughput quantification of fat cell size and number. Photomicrographs of H&E-stained paraffin sections of murine gonadal adipose were transformed using standard image processing/analysis algorithms to reduce background and enhance edge-detection. This allowed the isolation of individual adipocytes from which their area could be calculated. Performance was compared with manual measurements made from the same images, in which adipocyte area was calculated from estimates of the major and minor axes of individual adipocytes. Both methods identified an increase in mean adipocyte size in a murine model of obesity, with good concordance, although the calculation used to identify cell area from manual measurements was found to consistently over-estimate cell size. Here we report an accurate method to determine adipocyte area in histological sections that provides a considerable time saving over manual methods.

A topology-based score for pathway enrichment.

Investigators require intuitive tools to rationalize complex datasets generated by transcriptional profiling experiments. Pathway analysis methods, in which differentially expressed genes are mapped to databases of reference pathways to facilitate assessment of relative enrichment, lead investigators more effectively to biologically testable hypotheses. However, once a set of differentially expressed genes is isolated, pathway analysis approaches tend to ignore rich gene expression information and, moreover, do not exploit relationships between transcripts. In this article, we report the development of a new method in which both pathway topology and the magnitude of gene expression changes inform the scoring system, thereby providing a powerful filter in the enrichment of biologically relevant information. When four sample datasets were evaluated with this method, literature mining confirmed that those pathways germane to the physiological process under investigation were highlighted by our method relative to z-score overrepresentation calculations. Moreover, non-relevant processes were downgraded using the method described herein. The inclusion of expression and topological data in the calculation of a pathway regulation score (PRS) facilitated discrimination of key processes in real biological datasets. Specifically, by combining fold-change data for those transcripts exceeding a significance threshold, and by taking into account the potential for altered gene expression to impact upon downstream transcription, one may readily identify those pathways most relevant to pathophysiological processes.