SEARCH · PubMed Health

Results for “multiscale clustering”

Explore indexed PubMed citations for clinical trials, systematic reviews and public health research. Read source abstracts and follow each citation to its original PubMed record.

Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.

At least 19 recordsLinked to original sources

PubMed2025

Unsupervised multiscale clustering of single-cell transcriptomes to identify hierarchical structures of cell subtypes.

BACKGROUND: Cell clustering is an essential step in uncovering cellular architectures in single-cell RNA sequencing (scRNA-seq) data. However, the existing cell clustering approaches are not well designed to dissect complex structures of cellular landscapes at a finer resolution. RESULTS: Here, we develop a multiscale clustering (MSC) approach to construct a sparse cell-cell correlation network for unsupervised identification of de novo cell types and subtypes across multiple resolutions. Based upon simulated silver- and gold-standard data as well as real scRNA-seq data in diseases, MSC demonstrates significantly improved performance compared to established benchmark methods and reveals a biologically meaningful cell hierarchy to facilitate the discovery of novel disease-associated cell subtypes and mechanisms. CONCLUSIONS: We present MSC as a new single-cell multiscale clustering framework as a powerful tool for advancing discoveries in disease-associated cell populations using single-cell sequencing data.

Single-Cell Analysis↗

PubMed2002

Multiscale clustering in granular surface flows.

We investigate steady granular surface flows in a rotating drum and demonstrate the existence of rigid clusters of grains embedded in the flowing layer. These clusters appear to be fractal and their size is power law distributed from the grain size scale up to the thickness of the flowing layer. The implications of the absence of a characteristic length scale on available theoretical models of dense granular flows are discussed. Finally, we suggest a possible explanation of the difference between velocity profiles observed in surface flows and in flows down a rough inclined plane.

Journal Article↗

PubMed2003

The wavelet response as a multiscale NDT method.

We analyze interfaces by using reflected waves in the framework of the wavelet transform. First, we introduce the wavelet transform as an efficient method to detect and characterize a discontinuity in the acoustical impedance profile of a material. Synthetic examples are shown for both an isolated reflector and multiscale clusters of nearby defects. In the second part of the paper we present the wavelet response method as a natural extension of the wavelet transform when the velocity profile to be analyzed can only be remotely probed by propagating wavelets through the medium (instead of being directly convolved as in the wavelet transform). The wavelet response is constituted by the reflections of the incident wavelets on the discontinuities and we show that both transforms are equivalent when multiple scattering is neglected. We end this paper by experimentally applying the wavelet response in an acoustic tank to characterize planar reflectors with finite thicknesses.

Acoustics↗

PubMed2005

Modeling gene expression networks using fuzzy logic.

Gene regulatory networks model regulation in living organisms. Fuzzy logic can effectively model gene regulation and interaction to accurately reflect the underlying biology. A new multiscale fuzzy clustering method allows genes to interact between regulatory pathways and across different conditions at different levels of detail. Fuzzy cluster centers can be used to quickly discover causal relationships between groups of coregulated genes. Fuzzy measures weight expert knowledge and help quantify uncertainty about the functions of genes using annotations and the gene ontology database to confirm some of the interactions. The method is illustrated using gene expression data from an experiment on carbohydrate metabolism in the model plant Arabidopsis thaliana. Key gene regulatory relationships were evaluated using information from the gene ontology database. A new regulatory relationship concerning trehalose regulation of carbohydrate metabolism was also discovered in the extracted network.

Animals↗

PubMed2002

Automated 3-D echocardiography analysis compared with manual delineations and SPECT MUGA.

A major barrier for using 3-D echocardiography for quantitative analysis of heart function in routine clinical practice is the absence of accurate and robust segmentation and tracking methods necessary to make the analysis automatic. In this paper, we present an automated three-dimensional (3-D) echocardiographic acquisition and image-processing methodology for assessment of left ventricular (LV) function. We combine global image information provided by a novel multiscale fuzzy-clustering segmentation algorithm, with local boundaries obtained with phase-based acoustic feature detection. We then use the segmentation results to fit and track the LV endocardial surface using a 3-D continuous transformation. To our knowledge, this is the first report of a completely automated method. The protocol is evaluated in a small clinical case study (nine patients). We compare ejection fractions (EFs) computed with the new approach to those obtained using the standard clinical technique, single-photon emission computed tomography multigated acquisition. Errors on six datasets were found to be within six percentage points. A further two, with poor image quality, improved upon EFs from manually delineated contours, and the last failed due to artifacts in the data. Volume-time curves were derived and the results compared to those from manual segmentation. Improvement over an earlier published version of the method is noted.

Algorithms↗

PubMed2006

A multiscale expectation-maximization semisupervised classifier suitable for badly posed image classification.

This paper deals with the problem of badly posed image classification. Although underestimated in practice, bad-posedness is likely to affect many real-world image classification tasks, where reference samples are difficult to collect (e.g., in remote sensing (RS) image mapping) and/or spatial autocorrelation is relevant. In an image classification context affected by a lack of reference samples, an original inductive learning multiscale image classifier, termed multiscale semisupervised expectation maximization (MSEM), is proposed. The rationale behind MSEM is to combine useful complementary properties of two alternative data mapping procedures recently published outside of image processing literature, namely, the multiscale modified Pappas adaptive clustering (MPAC) algorithm and the sample-based semisupervised expectation maximization (SEM) classifier. To demonstrate its potential utility, MSEM is compared against nonstandard classifiers, such as MPAC, SEM and the single-scale contextual SEM (CSEM) classifier, besides against well-known standard classifiers in two RS image classification problems featuring few reference samples and modestly useful texture information. These experiments yield weak (subjective) but numerous quantitative map quality indexes that are consistent with both theoretical considerations and qualitative evaluations by expert photointerpreters. According to these quantitative results, MSEM is competitive in terms of overall image mapping performance at the cost of a computational overhead three to six times superior to that of its most interesting rival, SEM. More in general, our experiments confirm that, even if they rely on heavy class-conditional normal distribution assumptions that may not be true in many real-world problems (e.g., in highly textured images), semisupervised classifiers based on the iterative expectation maximization Gaussian mixture model solution can be very powerful in practice when: 1) there is a lack of reference samples with respect to the problem/model complexity and 2) texture information is considered negligible (i.e., a piecewise constant image model holds).

Algorithms↗

PubMed2000

Detection of microcalcifications by means of multiscale methods and statistical techniques.

The detection of clustered microcalcifications can help the radiologist to detect early breast cancer. Microcalcifications exhibit some important characteristics, such as small size and high luminosity. Use of a computer-aided diagnosis (CAD) method can prevent them being overlooked. In this report, a multiresolution analysis is performed based on a multilevel wavelet transformation. Decomposition produces sub-band images which become visible only as details of the different scales. Thereafter, all the images will be combined in a final image, in order to obtain an image that contains all the interest details at the scale where microcalcifications tend to appear. Once the image, called detail image, is obtained, it is necessary to determine which details correspond with microcalcifications. Statistical analysis of the histogram permits classification of the zones likely to contain microcalcifications. Applying this statistical techniques over the whole image and representing the results in a two-dimensional map, clustered microcalcification regions are clearly distinguishable.

Breast Neoplasms↗

PubMed2005

Object-based image analysis using multiscale connectivity.

This paper introduces a novel approach for image analysis based on the notion of multiscale connectivity. We use the proposed approach to design several novel tools for object-based image representation and analysis which exploit the connectivity structure of images in a multiscale fashion. More specifically, we propose a nonlinear pyramidal image representation scheme, which decomposes an image at different scales by means of multiscale grain filters. These filters gradually remove connected components from an image that fail to satisfy a given criterion. We also use the concept of multiscale connectivity to design a hierarchical data partitioning tool. We employ this tool to construct another image representation scheme, based on the concept of component trees, which organizes partitions of an image in a hierarchical multiscale fashion. In addition, we propose a geometrically-oriented hierarchical clustering algorithm which generalizes the classical single-linkage algorithm. Finally, we propose two object-based multiscale image summaries, reminiscent of the well-known (morphological) pattern spectrum, which can be useful in image analysis and image understanding applications.

Algorithms↗

PubMed2025

Characterizing the regulatory logic of transcriptional control at the DNA sequence level by ensembles of thermodynamic models.

MOTIVATION: Understanding how the genome encodes the regulatory logic of transcription is a main challenge of the post-genomic era, and can be overcome with the aid of customized computational tools. RESULTS: We report an automated framework for analyzing an ensemble of fits to data of a thermodynamics-based sequence-level model for transcriptional regulation. The fits are clustered accordingly with their intrinsic regulatory logic. A multiscale analysis enables visualization of quantitative features resulting from the deconvolution of the regulatory profile provided by multiple transcription factors interacting with the locus of a gene. Quantitative experimental data on reporters driven by the whole locus of the even-skipped gene in the blastoderm of Drosophila embryos was used for validating our approach. A few clusters of highly active DNA binding sites within the enhancers collectively modulate even-skipped gene transcription. Analysis of variable enhancers' length shows the importance of bound protein-protein interactions for transcriptional regulation. The interplay between activation and quenching enables function conservation of enhancers despite length variations. AVAILABILITY AND IMPLEMENTATION: The transcription factor level data used for performing the reported study is accessible in the input files in Zenodo and GitHub as well the full code. Additional data from formerly FlyEx database will be available under request.

Thermodynamics↗

PubMed2005

Functional magnetic resonance imaging activation detection: fuzzy cluster analysis in wavelet and multiwavelet domains.

PURPOSE: To present novel feature spaces, based on multiscale decompositions obtained by scalar wavelet and multiwavelet transforms, to remedy problems associated with high dimension of functional magnetic resonance imaging (fMRI) time series (when they are used directly in clustering algorithms) and their poor signal-to-noise ratio (SNR) that limits accurate classification of fMRI time series according to their activation contents. MATERIALS AND METHODS: Using randomization, the proposed method finds wavelet/multiwavelet coefficients that represent the activation content of fMRI time series and combines them to define new feature spaces. Using simulated and experimental fMRI data sets, the proposed feature spaces are compared to the cross-correlation (CC) feature space and their performances are evaluated. In these studies, the false positive detection rate is controlled using randomization. To compare different methods, several points of the receiver operating characteristics (ROC) curves, using simulated data, are estimated and compared. RESULTS: The proposed features suppress the effects of confounding signals and improve activation detection sensitivity. Experimental results show improved sensitivity and robustness of the proposed method compared to the conventional CC analysis. CONCLUSION: More accurate and sensitive activation detection can be achieved using the proposed feature spaces compared to CC feature space. Multiwavelet features show superior detection sensitivity compared to the scalar wavelet features.

Algorithms↗

PubMed2005

Longitudinal consistency of the relationship between depression symptoms and cognitive functioning in multiple sclerosis.

BACKGROUND: Lifetime prevalence rates of cognitive dysfunction and depression in multiple sclerosis (MS) have typically been reported to be approximately 50 percent. However, an inconsistent relationship between these two common features of MS has been reported in the literature. Because neurovegetative depression symptoms overlap with MS symptoms, it may be that literature inconsistencies can partly be explained by the fact that only those depression symptom clusters unambiguously reflective of depression are associated with cognitive dysfunction. OBJECTIVE: To explore the relationship between different depression symptom clusters and a battery of tests measuring cognitive domains commonly impaired in MS and was examined at two time points 3 years apart. METHODS: The Chicago Multiscale Depression Inventory was employed to measure mood, negative evaluative, and neurovegetative symptom clusters in 53 MS patients who were also administered a battery of neuropsychological tests. RESULTS: At time point 1, Mood and Evaluative Chicago Multiscale Depression Inventory scales were significantly associated with tasks of complex speeded attention, planning, and working memory. At time point 2, the Evaluative scale was still significantly associated with these domains, in addition to spatial memory; however, all of the significant correlations with the Mood scale dropped out. CONCLUSION: These results show that negative evaluative depression symptoms are most consistently predictive of cognitive dysfunction in MS. It may be that negative evaluative depression symptoms use up available cognitive capacity, thus compromising performance on cognitive capacity demanding tasks in MS patients.

Cognition Disorders↗

PubMed2006

Multiscale spatial Monte Carlo simulations: multigriding, computational singular perturbation, and hierarchical stochastic closures.

Monte Carlo (MC) simulation of most spatially distributed systems is plagued by several problems, namely, execution of one process at a time, large separation of time scales of various processes, and large length scales. Recently, a coarse-grained Monte Carlo (CGMC) method was introduced that can capture large length scales at reasonable computational times. An inherent assumption in this CGMC method revolves around a mean-field closure invoked in each coarse cell that is inaccurate for short-ranged interactions. Two new approaches are explored to improve upon this closure. The first employs the local quasichemical approximation, which is applicable to first nearest-neighbor interactions. The second, termed multiscale CGMC method, employs singular perturbation ideas on multiple grids to capture the entire cluster probability distribution function via short microscopic MC simulations on small, fine-grid lattices by taking advantage of the time scale separation of multiple processes. Computational strategies for coupling the fast process at small length scales (fine grid) with the slow processes at large length scales (coarse grid) are discussed. Finally, the binomial tau-leap method is combined with the multiscale CGMC method to execute multiple processes over the entire lattice and provide additional computational acceleration. Numerical simulations demonstrate that in the presence of fast diffusion and slow adsorption and desorption processes the two new approaches provide more accurate solutions in comparison to the previously introduced CGMC method.

Algorithms↗

PubMed2006

Multiscale detection of localized anomalous structure in aggregate disease incidence data.

We present a modelling framework for detection of potentially anomalous structure in aggregate spatial disease incidence data in a manner sensitive to localization at multiple scales and/or positions. The key technical contribution is the re-casting of the components of a multiscale disease mapping methodology, recently introduced by the authors in an earlier paper, into a form appropriate for hypothesis testing. In particular, we describe how hypotheses of spatially clustered variations in disease incidence may be linked in one-to-one correspondence with collections of hypotheses on the values of certain multiscale parameters associated with a user-defined hierarchy of nested partitions of an overall spatial region. A Bayesian hypothesis testing methodology is developed in the context of a standard Poisson measurement model, over the collection of possible multiscale hypotheses. We discuss the specification of hyper parameters and prior distributions on the space of models. The methodology is illustrated on both simulated and real data.

Cluster Analysis↗

PubMed2006

Laser induced hierarchical calcium phosphate structures.

The surface properties of biomedical implant materials control the dynamic interactions at tissue-implant interfaces. At such interfaces, if the nanoscale features influence protein interactions, those of the microscale and mesoscale aid cell orientation and provide tissue integration, respectively. It seems imperative that the synthetic materials expected to replace natural hard tissues are engineered to mimic the complexity of their hierarchical assembly. However, the current surface engineering approaches are single scaled. It is demonstrated that using laser surface engineering a controlled multiscale surface can be synthesized for bioactive functions. A systematic organization of bioactive calcium phosphate coating with multiphase composition on Ti-alloy substrate ranging from nano- to mesoscale has been achieved by effectively controlling the thermo physical interactions during laser processing. The morphology of the coating consisted of a periodic arrangement of Ti-rich and Ca-P-deficient star-like phases uniformly distributed inside a Ca-P-rich self-assembled cellular structure with the presence of CaO, alpha-tricalcium phosphate, CaTiO(3), TiO(2) and Ti phase in the coating matrix. The cellular structures ranged in diameter from 2.5 microm to 10 microm as an assembly of cuboid shaped particles of dimensions of approximately 200 nm x 1 microm. The multiscale texture also included nanoscale particles that are the precursors for many of these phases. The rapid cooling associated with the laser processing resulted in formation, organization and controlling dimensions of the Ca-P-rich glassy phase into a micron scale cellular morphology and submicron scale clusters of CaTiO(3) phase inside the cellular structures. The self-assembly of the coating into multiscale structure was influenced by chemical and physical interactions among the multiphases that evolved during laser processing.

Biocompatible Materials↗

PubMed2006

Unsupervised multiscale color image segmentation based on MDL principle.

We present an unsupervised multiscale color image segmentation algorithm. The basic idea is to apply mean shift clustering to obtain an over-segmentation and then merge regions at multiple scales to minimize the minimum description length criterion. The performance on the Berkeley segmentation benchmark campares favorably with some existing approaches.

Algorithms↗

PubMed2006

Multiscale hybrid linear models for lossy image representation.

In this paper, we introduce a simple and efficient representation for natural images. We view an image (in either the spatial domain or the wavelet domain) as a collection of vectors in a high-dimensional space. We then fit a piece-wise linear model (i.e., a union of affine subspaces) to the vectors at each downsampling scale. We call this a multiscale hybrid linear model for the image. The model can be effectively estimated via a new algebraic method known as generalized principal component analysis (GPCA). The hybrid and hierarchical structure of this model allows us to effectively extract and exploit multimodal correlations among the imagery data at different scales. It conceptually and computationally remedies limitations of many existing image representation methods that are based on either a fixed linear transformation (e.g., DCT, wavelets), or an adaptive uni-modal linear transformation (e.g., PCA), or a multimodal model that uses only cluster means (e.g., VQ). We will justify both quantitatively and experimentally why and how such a simple multiscale hybrid model is able to reduce simultaneously the model complexity and computational cost. Despite a small overhead of the model, our careful and extensive experimental results show that this new model gives more compact representations for a wide variety of natural images under a wide range of signal-to-noise ratios than many existing methods, including wavelets. We also briefly address how the same (hybrid linear) modeling paradigm can be extended to be potentially useful for other applications, such as image segmentation.

Algorithms↗

PubMed2006

Multiscale coarse graining of diblock copolymer self-assembly: from monomers to ordered micelles.

Starting from a microscopic lattice model, we investigate clustering, micellization, and micelle ordering in semidilute solutions of AB diblock copolymers in a selective solvent. To bridge the gap in length scales, from monomers to ordered micellar structures, we implement a two-step coarse-graining strategy, whereby the AB copolymers are mapped onto ultrasoft dumbells with monomer-averaged effective interactions between the centers of mass of the blocks. Monte Carlo simulations of this coarse-grained model yield clear-cut evidence for self-assembly into micelles with a mean aggregation number n approximately 100 beyond a critical concentration. At a slightly higher concentration the micelles spontaneously undergo a disorder-order transition to a cubic phase. We determine the effective potential between these micelles from first principles.

Journal Article↗

PubMed2002

Identification of ligand binding sites on proteins using a multi-scale approach.

Identification of a ligand binding site on a protein is pivotal to drug discovery. To date, no reliable and computationally feasible general approach to this problem has been published. Here we present an automated efficient method for determining binding sites on proteins for potential ligands without any a priori knowledge. Our method is based upon the multiscale concept where we deal with a hierarchy of models generated using a k-means clustering algorithm for the potential ligand. This is done in a simple approach whereby a potential ligand is represented by a growing number of feature points. At each increasing level of detail, a pruning of potential binding site is performed. A nonbonding energy function is used to score the interactions between molecules at each step. The technique was successfully employed to seven protein-ligand complexes. In the current paper we show that the algorithm considerably reduces the computational effort required to solve this problem. This approach offers real opportunities for exploiting the large number of structures that will evolve from structural genomics.

Algorithms↗