基于双模型融合的大熊猫头部图像分割

在大熊猫(Ailuropoda melanoleuca)的迁地保护和种群饲养管理中,及时、快速地进行个体识别和行为监测,对其健康管理具有至关重要的作用。圈养大熊猫健康状况通常由专门的饲养人员肉眼观测,人力成本高、效率低并且缺乏时效性。基于图像的动物个体识别与行为分析技术效率高、时间成本低,已经成为新的监测发展趋势。已有研究提出,通过大熊猫面部图像的检测和分析,可实现个体识别和行为分类。但该方法依然存在检测精度不足导致识别准确率难以提升的问题。本文提出一种基于YOLOv3和Mask R-CNN的双模型融合方法,实现了大熊猫头部图像分割和精准检测。包含3个部分：YOLOv3完成头部检测,Mask R-CNN完成大熊猫轮廓分割,然后将两个模型的输出进行交并比融合。结果显示,头部检测准确率为82.6%,大熊猫轮廓分割准确率为95.2%,总体头部轮廓分割准确率为87.1%。该方法对大熊猫头部图像的识别率和分割准确率高,为大熊猫的个体识别、性别分类提供了帮助,为行为分析提供了技术参考。     

Bayesian interval mapping of count trait loci based on zero-inflated generalized Poisson regression model

Count phenotypes with excessive zeros are often observed in the biological world. Researchers have studied many statistical methods for mapping the quantitative trait loci (QTLs) of zero-inflated count phenotypes. However, most of the existing methods consist of finding the approximate positions of the QTLs on the chromosome by genome-wide scanning. Additionally, most of the existing methods use the EM algorithm for parameter estimation. In this paper, we propose a Bayesian interval mapping scheme of QTLs for zero-inflated count data. The method takes advantage of a zero-inflated generalized Poisson (ZIGP) regression model to study the influence of QTLs on the zero-inflated count phenotype. The MCMC algorithm is used to estimate the effects and position parameters of QTLs. We use the Haldane map function to realize the conversion between recombination rate and map distance. Monte Carlo simulations are conducted to test the applicability and advantage of the proposed method. The effects of QTLs on the formation of mouse cholesterol gallstones were demonstrated by analyzing an mouse data set.     

A goodness-of-fit test for inhomogeneous spatial Poisson processes

We introduce a formal testing procedure to assess the fit ofan inhomogeneous spatial Poisson process model, based on a discrepancymeasure function thatis constructed from residuals obtained from the fitted model.We derive the asymptotic distributional properties of and develop a test statistic based on them.Our test statistic has a limiting standard normal distribution,so that the test can be performed by simply comparing the teststatistic with readily available critical values. We performa simulation study to assess the performance of the proposedmethod and apply it to a real data example.     

A 3D damage model for trabecular bone based on fabric tensors

Motivated by the role of damage in normal and pathological conditions of trabecular bone, a novel 3D constitutive law was developed that describes anisotropic elasticity and the rate-independent degradation in mechanical properties resulting from the growth of cracks or voids in the trabecular tissue. The theoretical model was formulated within the framework of continuum damage mechanics and based on two fabric tensors characterizing the local trabecular morphology. Experimental validation of the model was achieved by uniaxial and torsional testing of waisted bovine trabecular bone specimens. Strong correlations were found between cumulated permanent strain, reduction in elastic moduli and nonlinear postyield stress which support the hypothesis that these variables reflect the same underlying damage process.     

Identifying overrepresented concepts in gene lists from literature: a statistical approach based on Poisson mixture model

Background  

Large-scale genomic studies often identify large gene lists, for example, the genes sharing the same expression patterns. The interpretation of these gene lists is generally achieved by extracting concepts overrepresented in the gene lists. This analysis often depends on manual annotation of genes based on controlled vocabularies, in particular, Gene Ontology (GO). However, the annotation of genes is a labor-intensive process; and the vocabularies are generally incomplete, leaving some important biological domains inadequately covered.     

基于地理加权泊松模型的天然次生林进界株数空间分布与预测

基于帽儿山实验林场2004—2016年森林资源二类调查固定样地(共108块)数据,采用全局Poisson模型和4种空间尺度(2.5、5、10、15 km)下的地理加权泊松模型(geographically weighted Poisson regression, GWPR)对天然次生林进界株数的空间分布进行了研究,并对5种模型的拟合效果以及影响林分进界株数的因子进行了分析,利用莫兰指数描述了模型残差在全局和局域两种水平上的空间自相关性.结果表明: 本文所选的林分及地形因子都显著影响天然次生林进界株数的空间分布,林分平均胸径是最主要的影响因子;在小尺度(2.5 km)下GWPR模型拥有很高的拟合精度,产生了最大范围的模型参数估计值,得到了较好的模型参数局域化空间分布效果;在较小尺度(2.5和5 km)下GWPR模型产生了较小范围的模型残差,模型的稳定性得到提升;在小尺度(2.5 km)下GWPR模型残差的全局空间自相关性达到最低,局域空间自相关性显著减小,并形成了不同观测值少量聚类这一理想的空间分布模式;在对进界株数空间分布的模拟效果上,小尺度(2.5 km)下的局域模型明显好于全局模型.     

A note on robust inference from a conditional Poisson model

A randomised controlled trial to evaluate a training programme for physician-patient communication required the analysis of paired count data. The impact of departures from the Poisson assumption when paired count data are analysed through use of a conditional likelihood is illustrated. A simple approach to providing robust inference is outlined and illustrated.     

A simulation program based on a structured population model for biotechnological yeast processes

Summary A segregated population model for budding yeasts and a simulation program based on it are presented. They enable the study of bioprocesses utilizing yeasts in steady and perturbed conditions and in particular the comparison between the model predictions and the experimental results obtained by flow cytometry, which allows the measurement of segregated parameters of cell populations.Nomenclature a genealogical age - A parameter of the budding law - CV coefficient of variation - F _in(t) volumetric input flow - F _out(t) volumetric output flow - h parameter of the division law - K _s parameter of the Monod's law - m cell mass - M _i discretized cell mass - m _b (a,s) critical mass level for budding - m _p cell mass at the time of budding - n(t) cell number per unit volume - n _p number of sub-populations - n _c number of channels - p (a, i, j, k) discrete density function - Q parameter of the budding law - s(t) substrate concentration - S _in(t) substrate concentration in the input flow - t time - T _m minimal length of the budded phase - V(t) culture volume - x(t) biomass concentration - Y yield coefficient - channel width - (s) specific growth rate - _max parameter of the Monod's law     

Probability of disturbance of Poisson statistics in radioactive decay type processes

The phenomenon of regular dynamics of fine structure of sampling distributions in processes of radioactive decay type is discussed. The time dependence of the probability of their similarity indicates that the process measured is nonstationary. The most natural explanation for the disturbance of stationarity is the influence of gauging equipment, which in principle cannot be avoided completely. However, the universality of the effect and the presence of the correlation of the fine structure in independent synchronous processes leads one to suggest the existence of some global source. A mechanism of the generation of fine structure as a result of the influence of nontrivial topology of space on the decay process is proposed.     

A non-uniform bound on Poisson approximation in somatic cell hybrid model

The aim of this paper is a use of the Stein-Chen method to give a non-uniform bound for approximating the point probabilities of the number of pairs of chromosomes for which the Hamming distance is less than some fixed Hamming distance d by Poisson distribution.     

An enhanced version of a bone-remodelling model based on the continuum damage mechanics theory

The purpose of this work was to propose an enhancement of Doblaré and García's internal bone remodelling model based on the continuum damage mechanics (CDM) theory. In their paper, they stated that the evolution of the internal variables of the bone microstructure, and its incidence on the modification of the elastic constitutive parameters, may be formulated following the principles of CDM, although no actual damage was considered. The resorption and apposition criteria (similar to the damage criterion) were expressed in terms of a mechanical stimulus. However, the resorption criterion is lacking a dimensional consistency with the remodelling rate. We propose here an enhancement to this resorption criterion, insuring the dimensional consistency while retaining the physical properties of the original remodelling model. We then analyse the change in the resorption criterion hypersurface in the stress space for a two-dimensional (2D) analysis. We finally apply the new formulation to analyse the structural evolution of a 2D femur. This analysis gives results consistent with the original model but with a faster and more stable convergence rate.     

Bayesian semiparametric zero-inflated Poisson model for longitudinal count data

This paper presents new methods, using a Bayesian approach, for analyzing longitudinal count data with excess zeros and nonlinear effects of continuously valued covariates. In longitudinal count data there are many problems that can make the use of a zero-inflated Poisson (ZIP) model ineffective. These problems are unobserved heterogeneity and nonlinear effects of continuously valued covariates. Our proposed semiparametric model can simultaneously handle these problems in a unified framework. The framework accounts for heterogeneity by incorporating random effects and has two components. The parametric component of the model which deals with the linear effects of time invariant covariates and the non-parametric component which gives an arbitrary smooth function to model the effect of time or time-varying covariates on the logarithm of mean count. The proposed methods are illustrated by analyzing longitudinal count data on the assessment of an efficacy of pesticides in controlling the reproduction of whitefly.     

Metamodelling a 3D architectural root-system model to provide a simple model based on key processes and species functional groups

Plant and Soil - The architecture of root systems determines where and how much resources plants can extract from the soil, how they compete for soil resources, and how they interact with soil...     

Classification of breast lesions based on a dual S-shaped logistic model in dynamic contrast enhanced magnetic resonance imaging

This study proposes a novel dual S-shaped logistic model for automatically quantifying the characteristic kinetic curves of breast lesions and for distinguishing malignant from benign breast tumors on dynamic contrast enhanced (DCE) magnetic resonance (MR) images. D(α,β) is the diagnostic parameter derived from the logistic model. Significant differences were found in D(α,β) between the malignant benign groups. Fisher's Linear Discriminant analysis correctly classified more than 90% of the benign and malignant kinetic breast data using the derived diagnostic parameter (D(α,β)). Receiver operating characteristic curve analysis of the derived diagnostic parameter (D(α,β)) indicated high sensitivity and specificity to differentiate malignancy from benignancy. The dual S-shaped logistic model was effectively used to fit the kinetic curves of breast lesions in DCE-MR. Separation between benign and malignant breast lesions was achieved with sufficient accuracy by using the derived diagnostic parameter D(α,β) as the lesion's feature. The proposed method therefore has the potential for computer-aided diagnosis in breast tumors.     

A dual role of Cdk2 in DNA damage response

Once it was believed that Cdk2 was the master regulator of S phase entry. Gene knockout mouse studies of cell cycle regulators revealed that Cdk2 is dispensable for S phase initiation and progression whereby Cdk1 can compensate for the loss of Cdk2. Nevertheless, recent evidence indicates that Cdk2 is involved in cell cycle independent functions such as DNA damage repair. Whether these properties are unique to Cdk2 or also being compensated by other Cdks in the absence of Cdk2 is under extensive investigation. Here we review the emerging new role of Cdk2 in DNA damage repair and also discuss how the loss of Cdk2 impacts the G₁/S phase DNA damage checkpoint.