An Improved Object Detection Method for Mitosis Detection

Breast cancer grading is important for patient prognosis, and the mitosis count is one of the most important indicators for breast cancer grading. Traditional methods use handcraft features and deep learning based methods to detect mitosis in a classified model. These methods are time-consuming and difficult for practical clinical practice application. For this reason, this paper proposes an improved object detection method for automatic mitosis detection from histological images. First, we use a convolutional neural network (CNN) to automatically extract mitosis features. Then, we use the region proposed network (RPN) to locate a set of class-agnostic mitosis proposals. Finally, we use the improved R-CNN subnet to screen for mitosis from these proposals. Our approach achieved the best results in the ICPR2012 mitosis detection competition test dataset. Additionally, our proposed method is fast enough to be potentially used in clinical and health centers.

Nanoscale extracellular vesicle-derived DNA is superior to circulating cell-free DNA for mutation detection in early-stage non-small-cell lung cancer

Background

The comparison between relatively intact nanoscale extracellular vesicle-derived DNA (nEV-DNA) and fragmented circulating cell-free DNA (cfDNA) in mutation detection among patients with non-small-cell lung cancer (NSCLC) has not been carried out yet, and thus deserves investigation.

Patients and methods

Both nEV-DNA and cfDNA was obtained from 377 NSCLC patients with known EGFR mutation status and 69 controls. The respective EGFRE19del/T790M/L858R mutation status was interrogated with amplification-refractory-mutation-system-based PCR assays (ARMS-PCR).

Results

Neither nEV-DNA nor cfDNA levels show a strong correlation with tumor volumes. There is no correlation between cfDNA and nEV-DNA levels either. The detection sensitivity of nEV-DNA and cfDNA using ARMS-PCR in early-stage NSCLC was 25.7% and 14.2%, respectively, with 96.6% and 91.7% specificity, respectively. In late-stage NSCLC, both nEV-DNA and cfDNA show ∼80% sensitivity and over 95% specificity.

Conclusions

nEV-DNA is superior to cfDNA for mutation detection in early-stage NSCLC using ARMS-PCR. However, the advantages vanish in late-stage NSCLC.

Real-time parallel 3D multiple particle tracking with single molecule centrifugal force microscopy

Real-time tracking of multiple particles is key for quantitative analysis of dynamic biophysical processes and materials science via time-lapse microscopy image data, especially for single molecule biophysical techniques, such as magnetic tweezers and centrifugal force microscopy. However, real-time multiple particle tracking with high resolution is limited by the current imaging processes or tracking algorithms. Here, we demonstrate 1 nm resolution in three dimensions in real-time with a graphics-processing unit (GPU) based on a compute unified device architecture (CUDA) parallel computing framework instead of only a central processing unit (CPU). We also explore the trade-offs between processing speed and size of the utilized regions of interest and a maximum speedup of 137 is achieved with the GPU compared with the CPU. Moreover, we utilize this method with our recently self-built centrifugal force microscope (CFM) in experiments that track multiple DNA-tethered particles. Our approach paves the way for high-throughput single molecule techniques with high resolution and efficiency. LAY DESCRIPTION: Particles are widely used as probes in life sciences through their motions. In single molecule techniques such as optical tweezers and magnetic tweezers, microbeads are used to study intermolecular or intramolecular interactions via beads tracking. Also tracking multiple beads' motions could study cell-cell or cell-ECM interactions in traction force microscopy. Therefore, particle tracking is of key important during these researches. However, parallel 3D multiple particle tracking in real-time with high resolution is a challenge either due to the algorithm or the program. Here, we combine the performance of CPU and CUDA-based GPU to make a hybrid implementation for particle tracking. In this way, a speedup of 137 is obtained compared the program before only with CPU without loss of accuracy. Moreover, we improve and build a new centrifugal force microscope for multiple single molecule force spectroscopy research in parallel. Then we employed our program into centrifugal force microscope for DNA stretching study. Our results not only demonstrate the application of this program in single molecule techniques, also indicate the capability of multiple single molecule study with centrifugal force microscopy.

Cohort Profile: The Chongqing Cancer Cohort Study (CCCS) of the Urban Population in Southwest China

Purpose: Urbanization is causing an increasingly negative effect on public health in China. This study was established to examine the associations between socio-economic and environmental exposures and the potential impact of gene-environment interactions and cancer risk of urban population in Chongqing, China. Participants: The cohort was established in Beibei District of Chongqing in southwest China. Between March 2016 and December 2016, we enrolled 57,332 adults who were 40-69 years of age, and collected baseline data on demographic information, socio-economic status, lifestyle, family and personal disease histories through face-to-face interviews using a standardized questionnaire. Regular follow-up including face-to-face interviews will take place every 5 years. Findings to date: Ninety-nine percent (56658/57332) of the participants completed the baseline assessment. The eligible participants had a mean age of 54.8 years, and 51.42% were females. Nearly three-fifths of participants having a normal BMI (18.5 to 23.9 kg/m2) and one-third being overweight (24.0 to 27.9 kg/m2). Among males, 29.58% were smokers and 21.08% were alcohol users. Among females, 1.49% were smokers and 1.66% were alcohol users. Among all participants, 7.03% of males and 9.08% of females reported their family history of cancer. Future plans: The relationships of modifiable risk factors with the cancer risk will be analyzed. Meanwhile, participants will be closely tracked to minimize loss to follow-up. We plan to construct a risk prediction model on cancer and verify the prediction model by genome-wide association studies (GWAS). The successful completion of this cohort study will allow for better targeting of cancer screening to those at highest risk in urban population of China and provide clinicians and policymakers with a practical predication rule.

Nitrate-induced and in situ electrochemical activation synthesis of oxygen deficiencies-rich nickel/nickel (oxy)hydroxide hybrid films for enhanced electrocatalytic water splitting

Hydrogen produced by electrochemical water splitting offers a hopeful and renewable solution for addressing the global energy crisis; however, development of highly efficient non-noble-metal electrocatalysts remains a big challenge. Herein, we report a facile strategy to fabricate oxygen deficiencies-rich nickel/nickel (oxy)hydroxide hybrid films as efficient electrocatalysts for water splitting by in situ oxygen evolution reaction (OER) activation. Under OER conditions, the originally deposited Ni films from the ethaline-based deep eutectic solvent (DES) undergo a structural rearrangement with a phase transformation in the oxidation state from Ni(ii) to Ni(iii) at the surface. The change is coupled with an increase in oxygen deficiencies and a pronounced defective precursor is induced by the addition of nitrate ions, providing structural disordering and boosting the intrinsic activity of the catalyst, which strongly enhances the water splitting performance.

Parkinson's Disease Diagnosis via Joint Learning From Multiple Modalities and Relations

Parkinson's disease (PD) is a neurodegenerative progressive disease that mainly affects the motor systems of patients. To slow this disease deterioration, early and accurate diagnosis of PD is an effective way, which alleviates mental and physical sufferings by clinical intervention. In this paper, we propose a joint regression and classification framework for PD diagnosis via magnetic resonance and diffusion tensor imaging data. Specifically, we devise a unified multitask feature selection model to explore multiple relationships among features, samples, and clinical scores. We regress four clinical variables of depression, sleep, olfaction, cognition scores, as well as perform the classification of PD disease from the multimodal data. The multitask model explores the relationships at the level of clinical scores, image features, and subjects, to select the most informative and diseased-related features for diagnosis. The proposed method is evaluated on the public Parkinson's progression markers initiative dataset. The extensive experimental results show that the multitask framework can effectively boost the performance of regression and classification and outperforms other state-of-the-art methods. The computerized predictions of clinical scores and label for PD diagnosis may offer quantitative reference for decision support as well.

Protein-Protein Interactions Prediction via Multimodal Deep Polynomial Network and Regularized Extreme Learning Machine

Predicting the protein-protein interactions (PPIs) has played an important role in many applications. Hence, a novel computational method for PPIs prediction is highly desirable. PPIs endow with protein amino acid mutation rate and two physicochemical properties of protein (e.g., hydrophobicity and hydrophilicity). Deep polynomial network (DPN) is well-suited to integrate these modalities since it can represent any function on a finite sample dataset via the supervised deep learning algorithm. We propose a multimodal DPN (MDPN) algorithm to effectively integrate these modalities to enhance prediction performance. MDPN consists of a two-stage DPN, the first stage feeds multiple protein features into DPN encoding to obtain high-level feature representation while the second stage fuses and learns features by cascading three types of high-level features in the DPN encoding. We employ a regularized extreme learning machine to predict PPIs. The proposed method is tested on the public dataset of H. pylori, Human, and Yeast and achieves average accuracies of 97.87%, 99.90%, and 98.11%, respectively. The proposed method also achieves good accuracies on other datasets. Furthermore, we test our method on three kinds of PPI networks and obtain superior prediction results.

p.R180C mutation of glycosyltransferase B leads to B subgroup, an in vitro and in silico study

BACKGROUND AND OBJECTIVES

Dysfunctional glycosyltransferase A or B may lead to incomplete glycosylation of H antigen and atypical ABO blood group with weak A or B phenotypes, posing challenges for blood typing for transfusion.

MATERIALS AND METHODS

Serological studies and ABO gene analysis were performed. Flow cytometry was performed on HeLa cells transfected glycosyltransferase B expressing plasmids. Agglutination of transfected cells and total glycosyltransferase B transfer capacity were examined. Molecular dynamics simulations were used to explore possible dynamic conformational changes around the binding pocket.

RESULTS

We identified a mutation c.538C>T (p. R180C) of B allele in a Chinese donor and his father with ABw phenotype. In vitro expression study showed that mutation p.R180C, although not affecting expression of glycosyltransferase B, impaired H to B antigen conversion. The in silico analyses found that the residue Arg180 on the internal loop next to the entry of the binding pocket may have its long side chain salt-bridged with the highly flexible C-terminal carboxyl and contribute to the catalysis of H to B antigen conversion.

CONCLUSION

The p.R180C mutation impairs the conversion from H to B antigen and leads to weak B phenotype. Dynamic interaction between Arg180 and C-terminal of glycosyltransferase B may stabilize its binding with UDP-galactose and facilitate H/B antigen conversion.

Routes of transmission of influenza A H1N1, SARS CoV, and norovirus in air cabin: Comparative analyses

Identifying the exact transmission route(s) of infectious diseases in indoor environments is a crucial step in developing effective intervention strategies. In this study, we proposed a comparative analysis approach and built a model to simulate outbreaks of 3 different in-flight infections in a similar cabin environment, that is, influenza A H1N1, severe acute respiratory syndrome (SARS) coronavirus (CoV), and norovirus. The simulation results seemed to suggest that the close contact route was probably the most significant route (contributes 70%, 95% confidence interval [CI]: 67%-72%) in the in-flight transmission of influenza A H1N1 transmission; as a result, passengers within 2 rows of the index case had a significantly higher infection risk than others in the outbreak (relative risk [RR]: 13.4, 95% CI: 1.5-121.2, P = .019). For SARS CoV, the airborne, close contact, and fomite routes contributed 21% (95% CI: 19%-23%), 29% (95% CI: 27%-31%), and 50% (95% CI: 48%-53%), respectively. For norovirus, the simulation results suggested that the fomite route played the dominant role (contributes 85%, 95% CI: 83%-87%) in most cases; as a result, passengers in aisle seats had a significantly higher infection risk than others (RR: 9.5, 95% CI: 1.2-77.4, P = .022). This work highlighted a method for using observed outbreak data to analyze the roles of different infection transmission routes.

Joint regression and classification via relational regularization for Parkinson's disease diagnosis

It is known that the symptoms of Parkinson’s disease (PD) progress successively, early and accurate diagnosis of the disease is of great importance, which slows the disease deterioration further and alleviates mental and physical suffering. In this paper, we propose a joint regression and classification scheme for PD diagnosis using baseline multi-modal neuroimaging data. Specifically, we devise a new feature selection method via relational learning in a unified multi-task feature selection model. Three kinds of relationships (e.g., relationships among features, responses, and subjects) are integrated to represent the similarities among features, responses, and subjects. Our proposed method exploits five regression variables (depression, sleep, olfaction, cognition scores and a clinical label) to jointly select the most discriminative features for clinical scores prediction and class label identification. Extensive experiments are conducted to demonstrate the effectiveness of the proposed method on the Parkinson’s Progression Markers Initiative (PPMI) dataset. Our experimental results demonstrate that multi-modal data can effectively enhance the performance in class label identification compared with single modal data. Our proposed method can greatly improve the performance in clinical scores prediction and outperforms the state-of-art methods as well. The identified brain regions can be recognized for further medical analysis and diagnosis.

Sparse feature learning for multi-class Parkinson's disease classification

This paper solves the multi-class classification problem for Parkinson's disease (PD) analysis by a sparse discriminative feature selection framework. Specifically, we propose a framework to construct a least square regression model based on the Fisher's linear discriminant analysis (LDA) and locality preserving projection (LPP). This framework utilizes the global and local information to select the most relevant and discriminative features to boost classification performance. Differing in previous methods for binary classification, we perform a multi-class classification for PD diagnosis. Our proposed method is evaluated on the public available Parkinson's progression markers initiative (PPMI) datasets. Extensive experimental results indicate that our proposed method identifies highly suitable regions for further PD analysis and diagnosis and outperforms state-of-the-art methods.

Uncertainty in the response of terrestrial carbon sink to environmental drivers undermines carbon-climate feedback predictions

Terrestrial ecosystems play a vital role in regulating the accumulation of carbon (C) in the atmosphere. Understanding the factors controlling land C uptake is critical for reducing uncertainties in projections of future climate. The relative importance of changing climate, rising atmospheric CO2, and other factors, however, remains unclear despite decades of research. Here, we use an ensemble of land models to show that models disagree on the primary driver of cumulative C uptake for 85% of vegetated land area. Disagreement is largest in model sensitivity to rising atmospheric CO2 which shows almost twice the variability in cumulative land uptake since 1901 (1 s.d. of 212.8 PgC vs. 138.5 PgC, respectively). We find that variability in CO2 and temperature sensitivity is attributable, in part, to their compensatory effects on C uptake, whereby comparable estimates of C uptake can arise by invoking different sensitivities to key environmental conditions. Conversely, divergent estimates of C uptake can occur despite being based on the same environmental sensitivities. Together, these findings imply an important limitation to the predictability of C cycling and climate under unprecedented environmental conditions. We suggest that the carbon modeling community prioritize a probabilistic multi-model approach to generate more robust C cycle projections.

Immunotolerant indoleamine-2,3-dioxygenase is increased in condyloma acuminata

BACKGROUND

The tryptophan-depleting enzyme indoleamine-2,3-dioxygenase (IDO) is critical for the regulation of immunotolerance and plays an important role in immune-associated skin diseases.

OBJECTIVES

To analyse the level of IDO in condyloma acuminata (CA) and its role in this condition.

METHODS

IDO expression was assessed in the skin and peripheral blood of healthy controls and patients with CA. To assess the role of skin IDO in immunity, the ability of isolated epidermal cells to metabolize tryptophan and the influence on polyclonal T-cell mitogen (PHA)-stimulated T-cell proliferation were explored.

RESULTS

IDO median fluorescence intensities in peripheral blood mononuclear cells from patients with CA were similar to those from healthy controls. Immunohistochemistry showed that IDO⁺ cells were rare in normal skin and the control skin of patients with CA, but were greatly accumulated in wart tissue. Most fluorescence signals of IDO⁺ cells did not overlap with those of CD1a⁺ Langerhans cells. Human papillomavirus (HPV) DNA probe in situ hybridization showed a large number of IDO⁺ cells in the HPV^- site. Keratinocytes in the skin of healthy controls and the circumcised skin of patients with CA could minimally transform tryptophan into kynurenine, but IDO-competent epidermal cells from warts could transform tryptophan. In addition, these IDO-competent epidermal cells could inhibit PHA-stimulated T-cell proliferation. The addition of an IDO inhibitor, 1-methyl-d-tryptophan, restored the inhibited T-cell proliferation.

CONCLUSIONS

Abnormally localized high IDO expression might be involved in the formation of a local immunotolerant microenvironment.

Hierarchical Saliency Detection via Probabilistic Object Boundaries

Though there are many computational models proposed for saliency detection, few of them take object boundary information into account. This paper presents a hierarchical saliency detection model incorporating probabilistic object boundaries, which is based on the observation that salient objects are generally surrounded by explicit boundaries and show contrast with their surroundings. We perform adaptive thresholding operation on ultrametric contour map, which leads to hierarchical image segmentations, and compute the saliency map for each layer based on the proposed robust center bias, border bias, color dissimilarity and spatial coherence measures. After a linear weighted combination of multi-layer saliency maps, and Bayesian enhancement procedure, the final saliency map is obtained. Extensive experimental results on three challenging benchmark datasets demonstrate that the proposed model outperforms eight state-of-the-art saliency detection models.