BACKGROUND: Single-cell RNA sequencing (scRNA-seq) is a powerful tool for investigating cell abundance changes during tissue regeneration and remodeling processes. Differential cell abundance supports the initial clustering of all cells; then, the number of cells per cluster and sample are evaluated, and the dependence of these counts concerning the phenotypic covariates of the samples is studied. Analysis heavily depends on the clustering method. Partitioning Around Medoids (PAM or k-medoids) represents a well-established clustering procedure that leverages the downstream interpretation of clusters by pinpointing real individuals in the dataset as cluster centers (medoids) without reducing dimensions. Of note, PAM suffers from high computational costs and memory requirements. RESULTS: This paper proposes a method for differential abundance analysis using PAM as a clustering method and negative binomial regression as a statistical model to relate covariates to cluster/cell counts. We used this approach to study the differential cell abundance of human endometrial cell types throughout the natural secretory phase of the menstrual cycle. We developed a new R package -scellpam-, that incorporates an efficient parallel C++ implementation of PAM, and applied this package in this study. We compared the PAM-BS clustering method with other methods and evaluated both the computational aspects of its implementation and the quality of the classifications obtained using distinct published datasets with known subpopulations that demonstrate promising results. CONCLUSIONS: The implementation of PAM-BS, included in the scellpam package, exhibits robust performance in terms of speed and memory usage compared to other related methods. PAM allowed quick and robust clustering of sets of cells with a size ranging from 70,000 to 300,000 cells. https://cran.r-project.org/web/packages/scellpam/index.html . Finally, our approach provides important new insights into the transient subpopulations associated with the fertile time frame when applied to the study of changes in the human endometrium during the secretory phase of the menstrual cycle.

This paper describes a procedure to deal with large image collections obtained by smart city services based on interaction with citizens providing pictures. The semantic gap between the low-level image features and represented concepts and situations has been addressed using image retrieval techniques. A relevance feedback procedure is proposed for Content-Based Image Retrieval (CBIR) based on the modelling of user responses. One of the novelties of the proposal is that the feedback learning procedure can use the information that citizens themselves can provide when using these services.

Analyzing the morphological appearance and the spatial distribution of large dense-core vesicles (granules) in the cell cytoplasm is central to the understanding of regulated exocytosis. This paper is concerned with the automatic detection of granules and the statistical analysis of their spatial locations in different cell groups. We model the locations of granules of a given cell as a realization of a finite spatial point process and the point patterns associated with the cell groups as replicated point patterns of different spatial point processes. First, an algorithm to segment the granules using electron microscopy images is proposed. Second, the relative locations of the granules with respect to the plasma membrane are characterized by two functional descriptors: the empirical cumulative distribution function of the distances from the granules to the plasma membrane and the density of granules within a given distance to the plasma membrane. The descriptors of the different cells for each group are compared using bootstrap procedures. Our results show that these descriptors and the testing procedure allow discriminating between control and treated cells. The application of these novel tools to studies of secretion should help in the analysis of diseases associated with dysfunctional secretion, such as diabetes.

Analyzing spatio-temporal dependencies between different types of events is highly relevant to numerous biological phenomena (e.g. signalling and trafficking) especially as advances in probes and microscopy have facilitated imaging of dynamic processes in living cells. For many types of events, the segmented areas can overlap spatially and temporally forming random clumps. In this paper, we model binary image sequences of two different event types as a realization of a bivariate temporal random set and propose a non-parametric approach to quantify spatial and spatio-temporal interrelations using the pair-correlation, cross-covariance and the Ripley IK functions. Based on these summary statistics we propose a randomization procedure to test independence between event types by applying random toroidal shifts and Monte Carlo tests. A simulation study assessed the performance of the proposed estimators and showed that these statistics capture the spatio-temporal dependencies accurately. The estimation of the spatio-temporal interval of interactions was also obtained. The method was successfully applied to analyze the interdependencies of several endocytic proteins using image sequences of living cells and validated the procedure as a new way to automatically quantify dependencies between proteins in a formal and robust manner.

Total Internal Reflection Fluorescence Microscopy (TIRFM) greatly facilitates to imaging the first steps of endocytosis, a process whereby cells traffic cargo from the cell surface to endosomes. Using TIRFM, fluorescent-tagged endocytic proteins are observed as overlapped areas forming random clumps of different sizes, shapes and durations. A common procedure to segment these objects consists of thresholding the original gray-level images to produce binary sequences in which a pixel is covered or not by a given fluorescent-tagged protein. This binary logic is not appropriate because it leaves a free tuning parameter to be set by the user which can influence on the conclusions of the statistical analysis. Instead, we have adopted a more realistic approach, in which segmented binary images are modelled as a fuzzy temporal random set. Here, we propose some measures of spatio-temporal interactions based on the fuzzy counterparts of the pair-correlation function and the Ripley K-function. We used a randomization procedure to test for independence. Our results show that this procedure will permit biologists to examine and quantify the interactions between endocytic proteins robustly.

Counting spatially and temporally overlapping events in image sequences and estimating their shape-size and duration features are important issues in some applications. We propose a stochastic model, a particular case of the nonisotropic 3D Boolean model, for performing this analysis: the temporal Boolean model. Some probabilistic properties are derived and a methodology for parameter estimation from time-lapse image sequences is proposed using an explicit treatment of the temporal dimension. We estimate the mean number of germs per unit area and time, the mean grain size and the duration distribution. A wide simulation study in order to assess the proposed estimators showed promising results. The model was applied on biological image sequences of in-vivo cells in order to estimate new parameters such as the mean number and duration distribution of endocytic events. Our results show that the proposed temporal Boolean model is effective for obtaining information about dynamic processes which exhibit short-lived, but spatially and temporally overlapping events.

Exocytosis is an essential cellular trafficking process integral to the proper distribution and function of a plethora of molecules, including transporters, receptors, and enzymes. Moreover, incorrect protein targeting can lead to pathological conditions. Recently, the application of evanescent wave microscopy has allowed us to image the final steps of exocytosis. However, spatio-temporal analysis of fusion of constitutive vesicular traffic with the plasma membrane has not been systematically performed. Also, the spatial sites and times of vesicle fusion have not yet been analyzed together. In addition, more formal tests are required in testing biological hypotheses, rather than visual inspection combined with statistical descriptives. Ripley K-functions are used to examine the joint and marginal behavior of locations and fusion times. Semiautomatic detection and mapping of constitutive fusion sites reveals spatial and temporal clustering, but no dependency between the locations and times of fusion events. Our novel approach could be translated to other studies of membrane trafficking in health and diseases such as diabetes.

Image segmentation is a major problem in image processing, particularly in medical image analysis. A great number of segmentation procedures produce intermediate gray-scale images that can be understood as fuzzy sets. Additionally, some segmentation procedures tend to leave free tuning parameters (very influential in the final binary image) for the user. These different binary images can be easily aggregated (into a fuzzy set) by making use of fuzzy set theory. In any case, a single binary image is required so our interest is to associate a crisp set to a given fuzzy set in an intelligent and unsupervised manner. The main idea of this paper is to define the averages of a given fuzzy set by using different definitions of the mean of a random compact set. In particular, the average distance of Baddeley-Molchanov and the mean of Vorob'ev have been used. A theoretical study of some new definitions of fuzzy set averages has been performed. In particular, these averages have been obtained for L-R fuzzy numbers. Finally, we present a medical image application, that of retinal vessel detection. Some recent segmentation procedures have been revisited and modified using these new averages. The experimental results are very promising.

This paper presents a method for detecting abnormalities in spatial arrangements of cells within any tissue that can be described by different sets of relevant points. The method has been applied to the detection of subtle abnormalities in corneal endothelia. Images of this type of tissue can be characterized by two types of points: cell centroids and triple points associated with the apical intersections as it was proposed by Díaz.7 Both types of points jointly considered are modeled using a bivariate spatial point process; then a statistical analysis based on certain distributional descriptors proposed by Doguwa4,9 is carried out to discriminate severe and subtle abnormalities from controls. According to the results, the method constitutes a valid alternative to currently used techniques, which involve statistical descriptions of cell density, cell area and hexagonality, that fail in the detection of these subtle abnormalities.

Texture classification of an image or subimage is an important problem in texture analysis. Many procedures have been proposed. A global framework for texture classification based on random closed set theory is proposed in this paper. In this approach, a binary texture is considered as an outcome of a random closed set. Some distributional descriptors of this stochastic model are used as texture features in order to classify the binary texture, in particular spherical and linear contact distributions and K-functions. If a grayscale texture has to be classified, then the original texture is reduced to a multivariate random closed set where each component (a different random set) corresponds with those pixels verifying a local property. Again, some functional descriptors of the multivariate random closed set defined from the texture can be used as texture features to describe and classify the grayscale texture. Marginal and cross spherical and linear contact distributions and K-functions have been used. Experimental validation is provided by using Brodatz's database and another standard texture database.

This paper proposes new descriptors for binary and gray-scale images based on newly defined spatial size distributions (SSD). The main idea consists of combining a granulometric analysis of the image with a comparison between the geometric covariograms for binary images or the auto-correlation function for gray-scale images of the original image and its granulometric transformation; the usual granulometric size distribution then arises as a particular case of this formulation. Examples are given to show that in those cases in which a finer description of the image is required, the more complex descriptors generated from the SSD could be advantageously used. It is also shown that the new descriptors are probability distributions so their intuitive interpretation and properties can be appropriately studied from the probabilistic point of view. The usefulness of these descriptors in shape analysis is illustrated by some synthetic examples and their use in texture analysis is studied. Various cases of SSD and several former methods for texture classification are compared.

This article describes a method to detect and characterise the independent movement of several rigid objects in a sequence of images. No model is assumed for the moving objects, meanwhile a linear model in the image plane is accepted for the pixels of each object, which can be changed without losing the generality. The approach works on each pair of images. Relevant points are detected on them and then each moving object is identified as two clusters of similar relevant points, one per image, that perform a consistent movement. The consistent movement constraint is incorporated into the cluster building phase, which is different from previous approaches. The main property of the proposed method is its robustness, since minimal assumptions are assumed. Results are provided with tests on two different kinds of scenes: a laboratory one and a natural traffic scene.