An automatic method for removing empty camera trap images using ensemble learning

Camera traps often produce massive images, and empty images that do not contain animals are usually overwhelming. Deep learning is a machine‐learning algorithm and widely used to identify empty camera trap images automatically. Existing methods with high accuracy are based on millions of training samples (images) and require a lot of time and personnel costs to label the training samples manually. Reducing the number of training samples can save the cost of manually labeling images. However, the deep learning models based on a small dataset produce a large omission error of animal images that many animal images tend to be identified as empty images, which may lead to loss of the opportunities of discovering and observing species. Therefore, it is still a challenge to build the DCNN model with small errors on a small dataset. Using deep convolutional neural networks and a small‐size dataset, we proposed an ensemble learning approach based on conservative strategies to identify and remove empty images automatically. Furthermore, we proposed three automatic identifying schemes of empty images for users who accept different omission errors of animal images. Our experimental results showed that these three schemes automatically identified and removed 50.78%, 58.48%, and 77.51% of the empty images in the dataset when the omission errors were 0.70%, 1.13%, and 2.54%, respectively. The analysis showed that using our scheme to automatically identify empty images did not omit species information. It only slightly changed the frequency of species occurrence. When only a small dataset was available, our approach provided an alternative to users to automatically identify and remove empty images, which can significantly reduce the time and personnel costs required to manually remove empty images. The cost savings were comparable to the percentage of empty images removed by models.     

上海口岸2014-2015年入境人员肺结核筛查情况

为了解上海口岸入境人员肺结核的筛查情况及后续处理,防止结核病通过口岸跨境传播,本研究于2014年1月—2015年12月对所有在上海口岸办理入境体检的14岁以上人员进行结核病筛查,通过病史、体格检查和胸部X线摄影筛查疑似肺结核患者;对疑似肺结核患者进行痰细菌学检测、T‐SPOT ．TB和 Xpert MTB／RIF检测。结果显示,2014—2015年上海口岸入境人员共检出疑似肺结核患者215例,总检出率为229．76／10万;确诊肺结核患者33例,总检出率为35．27／10万,确诊率为15．3％。对210例疑似肺结核患者进行痰细菌学检测,结果显示结核分枝杆菌培阳率为14．3％,非结核分枝杆菌培阳率为17．1％。有95例和78例疑似肺结核患者分别接受 T‐SPOT ．TB和 Xpert MTB／RIF 检测,以痰细菌学检测为“金标准”,T‐SPOT ．TB的灵敏度为100％,特异度为49．4％;Xpert MTB／RIF的灵敏度为87．5％,特异度为96．8％。33例确诊肺结核患者中,25例（75．8％）离境,15例（45．5％）在离境前接受抗结核治疗,8例（24．2％）失访。本研究显示,上海口岸入境人员中肺结核确诊率仍有待提高。筛查与诊断中,T‐SPOT ．TB具备较高灵敏度, Xpert MTB／RIF具备较高特异度,两种方法均有较高应用价值,两者联用可提高检出率,缩短检出时间。对确诊病例或未确诊的可疑病例应加强后续监管。     

Evaluation of Diagnostic Value of Mediastinum for Differentiation of Drug Sensitive,Multi and Extensively Drug Resistant Tuberculosis Using Chest X-Rays

Background and ObjectiveThe rise of Drug Resistant Tuberculosis (DR TB), particularly Multi DR (MDR), and Extensively DR (XDR) has reduced the rate of control of the disease. Computer aided diagnosis using Chest X-rays (CXRs) can help in mass screening and timely diagnosis of DR TB, which is essential to administer proper treatment regimens. In CXRs, lungs and mediastinum are two significant regions which contain the information about the likelihood of DR TB. The objective of this work is to analyze the shape characteristics of lungs and mediastinum to improve the diagnostics accuracy for differentiation of Drug Sensitive (DS), MDR and XDR TB using computer aided diagnostics system.MethodsThe CXR images of DS and DR TB patients are obtained from a public database. The lung fields are segmented from the CXRs using Reaction Diffusion Level Set Evolution. Mediastinum is segmented from the delineated lung masks using Chan Vese model. The shape features from each lung and mediastinum masks are extracted and analysed. The discriminative power of individual and combination of both lung and mediastinum features are evaluated using machine learning techniques for classification of DS vs MDR, MDR vs XDR and DS vs XDR TB images. The performances of classifiers are compared using standard metrics.ResultsThe proposed segmentation methods are able to delineate lungs and mediastinum from the CXR images. The extracted lung and mediastinum features are found to be statistically significant (p < 0.05) for differentiation of DS and DR TB conditions. Using the combination of both lung and mediastinum features, Multi-Layer Perceptron classifier achieves maximum F-measure of 82.4%, 81.0% and 87.0% for differentiation of DS vs MDR, MDR vs XDR and DS vs XDR, respectively.ConclusionAnalysis of mediastinum along with the lungs in chest X-rays could improve the diagnostic performance for differentiation of drug sensitive and resistant TB conditions. The proposed methodology is able to differentiate DS, MDR and XDR TB, and found to be clinically relevant. Hence, this work is useful for computer-based early detection of DS and DR TB conditions.     

Wildlife surveillance using deep learning methods

相似文献   

miRDetect: A combinatorial approach for automated detection of novel miRNA precursors from plant EST data using homology and Random Forest classification

Identification of microRNAs from plants is a crucial step for understanding the mechanisms of pathways and regulation of genes. A number of tools have been developed for the detection of microRNAs from small RNA-seq data. However, there is a lack of pipeline for detection of miRNA from EST dataset even when a huge resource is publicly available and the method is known. Here we present miRDetect, a python implementation to detect novel miRNA precursors from plant EST data using homology and machine learning approach. 10-fold cross validation was applied to choose best classifier based on ROC, accuracy, MCC and F1-scores using 112 features. miRDetect achieved a classification accuracy of 93.35% on a Random Forest classifier and outperformed other precursor detection tools in terms of performance. The miRDetect pipeline aids in identifying novel plant precursors using a mixed approach and will be helpful to researchers with less informatics background.     

Healthy individuals that control a latent infection with Mycobacterium tuberculosis express high levels of Th1 cytokines and the IL-4 antagonist IL-4delta2

The majority of healthy individuals exposed to Mycobacterium tuberculosis will not develop disease and identifying what constitutes "protective immunity" is one of the holy grails of M. tuberculosis immunology. It is known that IFN-gamma is essential for protection, but it is also apparent that IFN-gamma levels alone do not explain the immunity/susceptibility dichotomy. The controversy regarding correlates of immunity persists because identifying infected but healthy individuals (those who are immune) has been problematic. We have therefore used recognition of the M. tuberculosis virulence factor early secretory antigenic target 6 to identify healthy, but infected individuals from tuberculosis (TB)-endemic and nonendemic regions (Ethiopia and Denmark) and have compared signals for cytokines expressed directly ex vivo with the pattern found in TB patients. We find that TB patients are characterized by decreased levels of Th1 cytokines and increased levels of IL-10 compared with the healthy infected and noninfected community controls. Interestingly, the healthy infected subjects exhibited a selective increase of message for the IL-4 antagonist, IL-4delta2, compared with both TB patients or noninfected individuals. These data suggest that long-term control of M. tuberculosis infection is associated not just with elevated Th1 responses but also with inhibition of the Th2 response.     

The prognosis of latent tuberculosis: can disease be predicted?

In humans, Mycobacterium tuberculosis persists for long periods in a clinically latent state, creating a huge reservoir of 'silent' tuberculosis (TB) (roughly one-third of the global population) from which new cases continually arise. A prognostic marker for active TB would enable targeted treatment of the small fraction of infected individuals who are most at risk of developing contagious TB, contributing greatly to TB control efforts. Here, we propose that TB-specific interferon-gamma release assays might be useful for identifying individuals with progressive infections who are likely to develop the disease. This might provide an unprecedented advantage for TB control, namely targeted preventive therapy for individuals who are most at risk of developing active contagious TB.     

Generalizable brain network markers of major depressive disorder across multiple imaging sites

Many studies have highlighted the difficulty inherent to the clinical application of fundamental neuroscience knowledge based on machine learning techniques. It is difficult to generalize machine learning brain markers to the data acquired from independent imaging sites, mainly due to large site differences in functional magnetic resonance imaging. We address the difficulty of finding a generalizable marker of major depressive disorder (MDD) that would distinguish patients from healthy controls based on resting-state functional connectivity patterns. For the discovery dataset with 713 participants from 4 imaging sites, we removed site differences using our recently developed harmonization method and developed a machine learning MDD classifier. The classifier achieved an approximately 70% generalization accuracy for an independent validation dataset with 521 participants from 5 different imaging sites. The successful generalization to a perfectly independent dataset acquired from multiple imaging sites is novel and ensures scientific reproducibility and clinical applicability.

Biomarkers for psychiatric disorders based on neuroimaging data have yet to be put to practical use. This study overcomes the problems of inter-site differences in fMRI data by using a novel harmonization method, thereby successfully constructing a generalizable brain network marker of major depressive disorder across multiple imaging sites.     

Latent M. tuberculosis infection--pathogenesis, diagnosis, treatment and prevention strategies

One third of the earths population is infected with Mycobacterium tuberculosis (Mtb), but only 5-10% of the infected individuals develop active tuberculosis (TB) over their lifetime. The remaining 90-95% stay healthy and are called latently infected individuals. They are the biggest reservoir of the tubercle bacilli and identifying the cases of latent TB is a part of the global plan of TB control. From the clinical point of view detection of latent TB infections (LTBI) in individuals with the highest active TB risk including cases of HIV infection, autoimmune inflammatory diseases or cancer, is a priority. This review summarizes the recent findings in the pathogenesis of latent TB, its diagnosis, treatment and prevention.     

Association of tuberculosis and polymorphisms in the promoter region of macrophage migration inhibitory factor (MIF) in a Southwestern China Han population

The macrophage migration inhibitory factor (MIF) is a pro-inflammatory cytokine that plays an important role in the pathogenesis of immune diseases. High levels of MIF have been detected in the sera of patients with tuberculosis (TB), and it has been proposed that MIF gene polymorphisms may influence the risk of developing TB. The aim of this study was to evaluate the potential relationship between functional polymorphisms of MIF and TB in a Han population from Southwestern China. TB patients (n=215) and healthy unrelated controls (n=245) were recruited for this study. Genomic DNA was isolated from all the participants. The MIF-173 G/C SNP was genotyped using polymerase chain reaction-restriction fragment length polymorphism (PCR-RFLP). The MIF-794 CATT(5-8) microsatellite was evaluated by direct sequencing of the subsequent PCR products. Association analysis of the two polymorphisms showed that the frequency of -173 (GC+CC) in TB patients and controls was 49.3% and 31.4%, respectively, which was statistically significant (OR=2.12, 95% CI=1.45-3.10, P<0.001); the frequencies of -794 (7/X+8/X) were 56.7% and 45.3%, respectively, also statistically significant between the TB and healthy controls (OR=1.58, 95% CI=1.10-2.29, P=0.015). In summary, Genetic variation in the MIF gene is closely associated with tuberculosis. Both the 173 (GC+CC) SNP and -794 (7/X+8/X) microsatellite increased the risk of Chinese Han developing TB.     

Differential MicroRNAs Expression in Serum of Patients with Lung Cancer, Pulmonary Tuberculosis, and Pneumonia

MicroRNAs (miRNAs) play critical regulatory roles in the physiological and pathological processes. The high stability of miRNAs in human serum represents attractive novel diagnostic biomarkers of clinical conditions. Several studies have shown that aberrant expression of miRNAs in human cancer including lung cancer, but little is known about their effects on some infectious lung diseases such as pulmonary tuberculosis (TB) and pneumonia. In this study, we investigated miRNA expression pattern in serum of Egyptian patients with lung cancer, TB, and pneumonia compared with matched healthy controls. Using microarray-based expression profiling followed by real-time quantitative polymerase chain reaction validation, we compared the levels of a series of circulating miRNAs (miR-21, miR-155, miR-182, and miR-197) in serum from patients with lung cancer (n = 65), pulmonary tuberculosis (n = 29), pneumonia (n = 29), and transudate (n = 16) compared with matched healthy controls (n = 37). MiRNA SNORD68 was the housekeeping endogenous control. We found that the serum levels of miR-21, miR-155, and miR-197 were significantly elevated in the patients with lung cancer and pneumonia whereas miR-182 and miR-197 levels were increased only in patients with lung cancer and TB, respectively, compared with controls. Receiver operating characteristic analysis revealed that miR-182, miR-155, and miR-197 have superior diagnostic potential in discriminating patients with lung cancer, pneumonia, and TB, respectively, from controls. Our results conclude that the differential expression of the four studied miRNAs can be potential non-invasive biomarkers for patients with lung cancer, TB and pneumonia.     

Exploring Radiologic Criteria for Glioma Grade Classification on the BraTS Dataset

ObjectivesGlioma grading using maching learning on magnetic resonance data is a growing topic. According to the World Health Organization (WHO), the classification of glioma discriminates between low grade gliomas (LGG), grades I, II; and high grade gliomas (HGG), grades III, IV, leading to major issues in oncology for therapeutic management of patients. A well-known dataset for machine-based grade prediction is the MICCAI Brain Tumor Segmentation (BraTS) dataset. However this dataset is not divided into WHO-defined LGG and HGG, since it combines grades I, II and III as “lower grades gliomas”, while its HGG category only presents grade IV glioblastoma multiform. In this paper we want to train a binary grade classifier and investigate the consistency of the original BraTS labels with radiologic criteria using machine-aided predictions.Material and methodsUsing WHO-based radiomic features, we trained a SVM classifier on the BraTS dataset, and used the prediction score histogram to investigate the behaviour of our classifier on the lower grade population. We also asked 5 expert radiologists to annotate BraTS images between low (as opposed to lower) grade and high grade glioma classes, resulting in a new groundtruth.ResultsOur first training reached 84.1% accuracy. The prediction score histogram allows us to identify the radiologically high grade patients among the original lower grade population of the BraTS dataset. Training another SVM on our new radiologically WHO-aligned groundtruth shows robust performances despite important class imbalance, reaching 82.4% accuracy.ConclusionOur results highlight the coherence of radiologic criteria for low grade versus high grade classification under WHO terms. We also show how the histogram of prediction scores and crossed prediction scores can be used as tools for data exploration and performance evaluation. Therefore, we propose to use our radiological groundtruth for future development on binary glioma grading.     

Biview Learning for Human Posture Segmentation from 3D Points Cloud

Posture segmentation plays an essential role in human motion analysis. The state-of-the-art method extracts sufficiently high-dimensional features from 3D depth images for each 3D point and learns an efficient body part classifier. However, high-dimensional features are memory-consuming and difficult to handle on large-scale training dataset. In this paper, we propose an efficient two-stage dimension reduction scheme, termed biview learning, to encode two independent views which are depth-difference features (DDF) and relative position features (RPF). Biview learning explores the complementary property of DDF and RPF, and uses two stages to learn a compact yet comprehensive low-dimensional feature space for posture segmentation. In the first stage, discriminative locality alignment (DLA) is applied to the high-dimensional DDF to learn a discriminative low-dimensional representation. In the second stage, canonical correlation analysis (CCA) is used to explore the complementary property of RPF and the dimensionality reduced DDF. Finally, we train a support vector machine (SVM) over the output of CCA. We carefully validate the effectiveness of DLA and CCA utilized in the two-stage scheme on our 3D human points cloud dataset. Experimental results show that the proposed biview learning scheme significantly outperforms the state-of-the-art method for human posture segmentation.     

High Prevalence of Pulmonary Tuberculosis but Low Sensitivity of Symptom Screening among HIV-Infected Pregnant Women in South Africa

Symptom screening is a recommended component of intensified case-finding (ICF) for pulmonary tuberculosis (TB) among HIV-infected individuals. Symptomatic individuals are further investigated to either exclude or diagnose pulmonary TB, thus reducing the number of individuals requiring costly laboratory investigation. Those with laboratory evaluations negative for pulmonary TB or who lack symptoms may be eligible for antiretroviral therapy (ART) and/or TB isoniazid preventive therapy (IPT). A four-part symptom screen has been recommended by the World Health Organization (WHO) for identifying TB suspects and those unlikely to have TB. A meta-analysis of studies among HIV-infected individuals calculated a sensitivity of 90.1% for the four-part symptoms screen - of any of cough, fever, night sweats, or weight loss - among patients in clinical care, making it an effective tool for identifying most patients with TB. An important population for intensified case-finding not included in that meta-analysis was HIV-infected pregnant women. We undertook a cross-sectional survey among HIV-infected pregnant women receiving prenatal care at community clinics in South Africa. We obtained a four-symptom review and sputum smear microscopy and mycobacterial culture on all participants. Among 1415 women, 226 (16%) had a positive symptom screen, and 35 (2.5%) were newly diagnosed with culture-positive TB. Twelve were on TB treatment at the time of screening, yielding 47 (3.3%) women with prevalent TB. Symptom screening among women without known TB had a sensitivity of 28% and specificity of 84%. The poor performance of symptom screening to identify women with TB suggests that other approaches may be needed for intensified case-finding to be effective for this population.     

Altered microRNA Signatures in Sputum of Patients with Active Pulmonary Tuberculosis

Role of microRNA (miRNA) has been highlighted in pathogen-host interactions recently. At present, their role in active pulmonary tuberculosis is unknown. The aim of the study was to delineate miRNA expression in sputum supernatant of patients with active pulmonary tuberculosis. Expression of miRNAs was evaluated by microarray analysis and differentially expressed miRNAs were validated by RT-qPCR. Secreted cytokines TNF-α and IL-6 were measured by ELISA. We found that 95 miRNAs were differentially expressed between tuberculosis group and controls. More miRNAs (52 out of 95 miRNAs) were underexpressed than overexpressed during tuberculosis infection. Overexpression of miR-3179, miR-147 and underexpression of miR-19b-2* in TB group compared with controls were confirmed in the validation cohort. TNF-α and IL-6 levels were not significantly altered between TB group and controls. For the first time, differential expression of miRNAs in sputum was found in active pulmonary tuberculosis. The study provides rationale for identifying the role of miRNAs in the pathogenesis of pulmonary tuberculosis and indicates potential for miRNA-based therapeutic strategies.     

Contribution of the P2X7 1513A/C loss-of-function polymorphism to extrapulmonary tuberculosis susceptibility in Tunisian populations

The P2X7 receptor has been found to be linked to an increased risk for tuberculosis in some populations. In this study, we investigate whether the P2X7 receptor plays a role in increasing susceptibility to tuberculosis in Tunisia. We examined two 1513A/C and -762T/C polymorphisms at the P2X7 receptor in 168 patients with pulmonary TB (pTB), 55 patients with extrapulmonary TB (epTB) and 150 blood donors from Tunisia. Genotyping of 1513A/C and -762T/C polymorphisms was performed in purified genomic DNA using PCR-restriction fragment length polymorphism and allele-specific PCR, respectively. The 1513C, CC and AC loss-of-function allele and genotypes were overrepresented in the epTB group compared with the control group (45% vs. 17%, P=10(-8) ; 24% vs. 4%, P=3 × 10(-7) ; 42% vs. 27%, P=10(-3) , respectively). Additionally, they were associated with 3.83-, 11.86- and 3.15-fold risks of developing this clinical tuberculosis form, respectively. No associations between the -762T/C polymorphism and tuberculosis disease, as well as disease anatomic location were observed. Collectively, our results suggest that the P2X7 1513A/C loss-of-function polymorphism may contribute to susceptibility to epTB in Tunisian populations.     

Glaucoma classification based on intra-class and extra-class discriminative correlation and consensus ensemble classifier

Automatic classification of glaucoma from fundus images is a vital diagnostic tool for Computer-Aided Diagnosis System (CAD). In this work, a novel fused feature extraction technique and ensemble classifier fusion is proposed for diagnosis of glaucoma. The proposed method comprises of three stages. Initially, the fundus images are subjected to preprocessing followed by feature extraction and feature fusion by Intra-Class and Extra-Class Discriminative Correlation Analysis (IEDCA). The feature fusion approach eliminates between-class correlation while retaining sufficient Feature Dimension (FD) for Correlation Analysis (CA). The fused features are then fed to the classifiers namely Support Vector Machine (SVM), Random Forest (RF) and K-Nearest Neighbor (KNN) for classification individually. Finally, Classifier fusion is also designed which combines the decision of the ensemble of classifiers based on Consensus-based Combining Method (CCM). CCM based Classifier fusion adjusts the weights iteratively after comparing the outputs of all the classifiers. The proposed fusion classifier provides a better improvement in accuracy and convergence when compared to the individual algorithms. A classification accuracy of 99.2% is accomplished by the two-level hybrid fusion approach. The method is evaluated on the public datasets High Resolution Fundus (HRF) and DRIVE datasets with cross dataset validation.     

Identification of potential lipid biomarkers for active pulmonary tuberculosis using ultra-high-performance liquid chromatography-tandem mass spectrometry

Early diagnosis of active pulmonary tuberculosis (TB) is the key to controlling the disease. Host lipids are nutrient sources for the metabolism of Mycobacterium tuberculosis. In this research work, we used ultra-high-performance liquid chromatography-tandem mass spectrometry to screen plasma lipids in TB patients, lung cancer patients, community-acquired pneumonia patients, and normal healthy controls. Principal component analysis, orthogonal partial least squares discriminant analysis, and K-means clustering algorithm analysis were used to identify lipids with differential abundance. A total of 22 differential lipids were filtered out among all subjects. The plasma phospholipid levels were decreased, while the cholesterol ester levels were increased in patients with TB. We speculate that the infection of M. tuberculosis may regulate the lipid metabolism of TB patients and may promote host-assisted bacterial degradation of phospholipids and accumulation of cholesterol esters. This may be related to the formation of lung cavities with caseous necrosis. The results of receiver operating characteristic curve analysis revealed four lipids such as phosphatidylcholine (PC, 12:0/22:2), PC (16:0/18:2), cholesteryl ester (20:3), and sphingomyelin (d18:0/18:1) as potential biomarkers for early diagnosis of TB. The diagnostic model was fitted by using logistic regression analysis and combining the above four lipids with a sensitivity of 92.9%, a specificity of 82.4%, and the area under the curve (AUC) value of 0.934 (95% CI 0.873 – 0.971). The machine learning method (10-fold cross-validation) demonstrated that the model had good accuracy (0.908 AUC, 85.3% sensitivity, and 85.9% specificity). The lipids identified in this study may serve as novel biomarkers in TB diagnosis. Our research may pave the foundation for understanding the pathogenesis of TB.     

Deep Learning Using Havrda-Charvat Entropy for Classification of Pulmonary Optical Endomicroscopy

1) ObjectivePulmonary optical endomicroscopy (POE) is an imaging technology in real time. It allows to examine pulmonary alveoli at a microscopic level. Acquired in clinical settings, a POE image sequence can have as much as 25% of the sequence being uninformative frames (i.e. pure-noise and motion artifacts). For future data analysis, these uninformative frames must be first removed from the sequence. Therefore, the objective of our work is to develop an automatic detection method of uninformative images in endomicroscopy images.2) Material and methodsWe propose to take the detection problem as a classification one. Considering advantages of deep learning methods, a classifier based on CNN (Convolutional Neural Network) is designed with a new loss function based on Havrda-Charvat entropy which is a parametrical generalization of the Shannon entropy. We propose to use this formula to get a better hold on all sorts of data since it provides a model more stable than the Shannon entropy.3) ResultsOur method is tested on one POE dataset including 3895 distinct images and is showing better results than using Shannon entropy and behaves better with regard to the problem of overfitting. We obtain 70% of accuracy with Shannon entropy versus 77 to 79% with Havrda-Charvat.4) ConclusionWe can conclude that Havrda-Charvat entropy is better suited for restricted and or noisy datasets due to its generalized nature. It is also more suitable for classification in endomicroscopy datasets.