Search PubMed⌕ Search

SEARCH · Search PubMed

Results for “Transformer model”

Search indexed PubMed citations on genomics, clinical trials, systematic reviews and public health. Explore titles, authors and supplied subject terms, then open the PubMed record.

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

At least 91 records · Page 5Linked to original sources
PubMed2001

A profile conditional likelihood approach for the semiparametric transformation regression model with missing covariates.

We propose a profile conditional likelihood approach to handle missing covariates in the general semiparametric transformation regression model. The method estimates the marginal survival function by the Kaplan-Meier estimator, and then estimates the parameters of the survival model and the covariate distribution from a conditional likelihood, substituting the Kaplan-Meier estimator for the marginal survival function in the conditional likelihood. This method is simpler than full maximum likelihood approaches, and yields consistent and asymptotically normally distributed estimator of the regression parameter when censoring is independent of the covariates. The estimator demonstrates very high relative efficiency in simulations. When compared with complete-case analysis, the proposed estimator can be more efficient when the missing data are missing completely at random and can correct bias when the missing data are missing at random. The potential application of the proposed method to the generalized probit model with missing continuous covariates is also outlined.

Humans↗
PubMed1999

Automatic retinal image registration scheme using global optimization techniques.

Retinal image registration is commonly required in order to combine the complementary information in different retinal modalities. In this paper, a new automatic scheme to register retinal images is presented and is currently tested in a clinical environment. The scheme considers the suitability and efficiency of different image transformation models and function optimization techniques, following an initial preprocessing stage. Three different transformation models--affine, bilinear and projective--as well as three optimization techniques--downhill simplex method, simulated annealing and genetic algorithms--are investigated and compared in terms of accuracy and efficiency. The registration of 26 pairs of Fluoroscein Angiography and Indocyanine Green Chorioangiography images with the corresponding Red-Free retinal images, showed the superiority of combining genetic algorithms with the affine and bilinear transformation models. A comparative study of the proposed automatic registration scheme against the manual method, commonly used in the clinical practice, is finally presented showing the advantage of the proposed automatic scheme in terms of accuracy and consistency.

Algorithms↗
PubMed2005

Spatio-temporal free-form registration of cardiac MR image sequences.

In this paper we present two registration algorithms for the spatio-temporal alignment of cardiac MR image sequences. Both algorithms have the ability to correct spatial misalignment between the images sequences caused by global and local shape differences. In addition, they have the ability to correct temporal misalignment caused by differences in the length of the cardiac cycles and by differences in the dynamic properties of the hearts. The algorithms use a 4D deformable transformation model which is separated into spatial and temporal components. The first registration algorithm optimizes the spatial and temporal transformation models simultaneously, while the second registration algorithm optimizes the temporal transformation component before optimizing the spatial component. For the evaluation of the spatio-temporal registration methods we have acquired 15 MR image sequences from healthy volunteers. The registration methods were quantitatively evaluated by measuring the overlap and surface distance of anatomical regions and qualitatively by visual inspection. The results demonstrate that a significant improvement in the alignment of the image sequences is achieved by the use of the deformable spatio-temporal transformation model. We demonstrate the use of the method for the construction of a probabilistic MR cardiac atlas representing the anatomy and function of a healthy heart.

Algorithms↗
PubMed2004

Investigation of three new mouse mammary tumor cell lines as models for transforming growth factor (TGF)-beta and Neu pathway signaling studies: identification of a novel model for TGF-beta-induced epithelial-to-mesenchymal transition.

INTRODUCTION: This report describes the isolation and characterization of three new murine mammary epithelial cell lines derived from mammary tumors from MMTV (mouse mammary tumor virus)/activated Neu + TbetaRII-AS (transforming growth factor [TGF]-beta type II receptor antisense RNA) bigenic mice (BRI-JM01 and BRI-JM05 cell lines) and MMTV/activated Neu transgenic mice (BRI-JM04 cell line). METHODS: The BRI-JM01, BRI-JM04, and BRI-JM05 cell lines were analyzed for transgene expression, their general growth characteristics, and their sensitivities to several growth factors from the epidermal growth factor (EGF) and TGF-beta families (recombinant human EGF, heregulin-beta1 and TGF-beta1). The BRI-JM01 cells were observed to undergo a striking morphologic change in response to TGF-beta1, and they were therefore further investigated for their ability to undergo a TGF-beta-induced epithelial-to-mesenchymal transition (EMT) using motility assays and immunofluorescence microscopy. RESULTS: We found that two of the three cell lines (BRI-JM04 and BRI-JM05) express the Neu transgene, whereas, unexpectedly, both of the cell lines that were established from MMTV/activated Neu + TbetaRII-AS bigenic tumors (BRI-JM01 and BRI-JM05) do not express the TbetaRII-AS transgene. The cuboidal BRI-JM01 cells exhibit a short doubling time and are able to form confluent monolayers. The BRI-JM04 and BRI-JM05 cell lines are morphologically much less uniform, grow at a much slower rate, and do not form confluent monolayers. Only the BRI-JM05 cells can form colonies in soft agar. In contrast, all three cell lines form colonies in Matrigel, although the BRI-JM04 and BRI-JM05 cell lines do so more efficiently than the BRI-JM01 cell line. All three cell lines express the cell surface marker E-cadherin, confirming their epithelial character. Proliferation assays showed that the three cell lines respond differently to recombinant human EGF and heregulin-beta1, and that all are growth inhibited by TGF-beta1, but that only the BRI-JM01 cell line undergoes an EMT and exhibits increased motility upon TGF-beta1 treatment. CONCLUSION: We suggest that the BRI-JM04 and BRI-JM05 cell lines can be used to investigate Neu oncogene driven mammary tumorigenesis, whereas the BRI-JM01 cell line will be useful for studying TGF-beta1-induced EMT.

Animals↗
PubMed1996

Linear systems analysis of functional magnetic resonance imaging in human V1.

The linear transform model of functional magnetic resonance imaging (fMRI) hypothesizes that fMRI responses are proportional to local average neural activity averaged over a period of time. This work reports results from three empirical tests that support this hypothesis. First, fMRI responses in human primary visual cortex (V1) depend separably on stimulus timing and stimulus contrast. Second, responses to long-duration stimuli can be predicted from responses to shorter duration stimuli. Third, the noise in the fMRI data is independent of stimulus contrast and temporal period. Although these tests can not prove the correctness of the linear transform model, they might have been used to reject the model. Because the linear transform model is consistent with our data, we proceeded to estimate the temporal fMRI impulse-response function and the underlying (presumably neural) contrast-response function of human V1.

Artifacts↗
PubMed1997

Producing vascularized bone by heterotopic bone induction and guided tissue regeneration: a silicone membrane-isolated latissimus dorsi island flap in a rat model.

Transformation of mesenchymal-type tissue into cartilage and bone can be induced by bone morphogenetic protein, and by its parent substratum, demineralized bone matrix. The authors were interested in transforming muscle island flaps into vascularized bone that could be used as autogeneic skeletal replacement parts. In Wistar rats, tubular latissimus dorsi muscle island flaps were created, using microsurgical techniques. The flaps were inserted by a cylinder of demineralized bone matrix (DBM) and enclosed in silicone rubber membrane tubes. The animals were followed-up for 10, 21, or 35 days. Rats with DBM implanted in muscle pouches served as controls. Quantitative radiomorphometry and qualitative histology were performed. A statistically significant linear time-related increase in radiomorphometrically-measured calcified tissue was found in the flaps with DBM from 10 days to 5 weeks. At 3 and 5 weeks, lamellar and cancellous bone with fully developed marrow was detected microscopically. There was no significant difference in bone quantity in the island flaps after 35 days, compared with the muscle pouches implanted with DBM, although the difference at 21 days was still significantly in favor of the island flaps. Using allogeneic DBM in rat muscle island flaps surrounded with a silicone membrane, it was possible to generate in vivo autogenous new bone with a good vascular supply and good mobility, allowing later transfer to another site. The experiment provided a basic technique that can be used as a standard in testing various osteoinductive substances for the production of vascular-pedicled new bone.

Animals↗
PubMed2005

Creating oral squamous cancer cells: a cellular model of oral-esophageal carcinogenesis.

Immortalization and malignant transformation are important steps in tumor development. The ability to induce these processes from normal human epithelial cells with genetic alterations frequently found in the corresponding human cancer would significantly enhance our understanding of tumor development. Alterations in several key intracellular regulatory pathways (the pRB, p53, and mitogenic signaling pathways and the telomere maintenance system) appear to be sufficient for the neoplastic transformation of normal human cells. Nevertheless, in vitro transformation models to date depend on viral oncogenes, most prominently the simian virus 40 early region, to induce immortalization and malignant transformation of normal human epithelial cells. Here, we demonstrate a transformation model creating oral-esophageal cancer cells by using a limited set of genetic alterations frequently observed in the corresponding human cancer. In a stepwise model, cyclin D1 overexpression and p53 inactivation led to immortalization of oral keratinocytes. Additional ectopic epithelial growth factor receptor overexpression followed by c-myc overexpression as well as consecutive reactivation of telomerase induced by epithelial growth factor receptor sufficed to transform oral epithelial cells, truly recapitulating the development of the corresponding human disease.

Carcinoma, Squamous Cell↗
PubMed1984

Tracheal epithelial cell transformation: a model system for studies on neoplastic progression.

Most in vitro transformation studies have been conducted with fibroblast cultures of various origins. The phenotypic changes known to accompany transformation are therefore primarily those that are typical for transformed fibroblasts. Little information exists concerning phenotypic changes occurring during transformation of epithelial cells in vitro. However, recently a number of transformation studies have been reported with tracheal epithelium as a prototype for epithelium from the conducting airways. The initial studies were carried out with organ culture-cell culture systems. These studies reported the qualitative phenotypic changes developing in primary outgrowth cultures after exposure to the direct acting carcinogen N-methyl-N'-nitrosoguanidine. The phenotypic changes observed are all related to changes of in vitro growth characteristics. Several stages can be observed as the cell cultures progress from a "carcinogen altered" to the neoplastic state. While these studies laid the groundwork for the epithelial transformation field, they did not permit quantitation of transformants, since the size of the exposed cell population is unknown. More recently transformation systems with dispersed primary tracheal epithelial cells have been developed which allow quantitation of transformed phenotypes. These systems are being used for clonal analysis of the process of epithelial cell transformation and to study progression and promotion during development of neoplastic transformation.

Animals↗
PubMed2006

The applicability of Bass's model of transformational, transactional, and laissez-faire leadership in the hospital administrative environment.

The purpose of this study is to evaluate empirically in the hospital administrative environment the relationship of leadership behaviors to subordinate manager's perceived outcomes, through examination of B. M. Bass's (1985) model of transformational, transactional, and laissez-faire leadership. The author measured leadership orientation and outcome factors through subordinate managers' ratings of hospital CEOs using a questionnaire, which asked: Is there a relationship between the leadership styles of hospital CEOs and subordinate managers' self-reported willingness to exert extra effort, perception of leader effectiveness and satisfaction with their leader? Findings revealed that the relationship between transformational leadership and the outcome factors were stronger and more positive than were the transactional and laissez-faire styles. These findings are consistent with the hierarchal patterns reported and support the universality of the model.

Data Collection↗
PubMed2000

A recurrent model of transformation invariance by association.

This paper describes an investigation of a recurrent artificial neural network which uses association to build transform-invariant representations. The simulation implements the analytic model of Parga and Rolls [(1998). Transform-invariant recognition by association in a recurrent network. Neural Computation 10(6), 1507-1525.] which defines multiple (e.g. "view") patterns to be within the basin of attraction of a shared (e.g. "object") representation. First, it was shown that the network could store and correctly retrieve an "object" representation from any one of the views which define that object, with capacity as predicted analytically. Second, new results extended the analysis by showing that correct object retrieval could occur where retrieval cues were distorted; where there was some association between the views of different objects; and where connectivity was diluted, even when this dilution was asymmetric. The simulations also extended the analysis by showing that the system could work well with sparse patterns; and showing how pattern sparseness interacts with the number of views of each object (as a result of the statistical properties of the pattern coding) to give predictable object retrieval performance. The results thus usefully extend a recurrent model of invariant pattern recognition.

Artificial Intelligence↗
PubMed2002

Phylogenetic placement of retropinnid fishes: data set incongruence can be reduced by using asymmetric character state transformation costs.

We used mitochondrial DNA sequences to determine the phylogenetic placement of southern smelts (Retropinnidae), a group of diadromous fishes endemic to New Zealand and Australia. Our genetic data strongly support a sister group relationship between retropinnids and northern hemisphere smelts (Osmeridae), a relationship that seems consistent with the similar appearance and life history strategies of these two groups. Our analysis indicates that Retropinnidae and Osmeridae together represent the sister group to the southern hemisphere galaxiid fishes (Galaxiidae). However, this finding conflicts with several recent osteological analyses, which supported a sister relationship for Retropinnidae and Galaxiidae, giving a monophyletic southern hemisphere assemblage (Galaxioidea). We review cases of incongruence and discuss factors that might explain significant disagreement between molecular and morphological data matrices. We suggest that repeated evolutionary simplification may have undermined the accuracy of morphological hypotheses of osmeroid relationships. Although equally weighted parsimony analysis of morphological data rejects the molecular hypothesis (Osmeridae + Retropinnidae), implementation of a range of weighting schemes suggests that incongruence is nonsignificant under asymmetric character transformation models. We propose that a simple "equal transformation cost" parsimony analysis may be biologically unrealistic, especially when reductive homoplasy is widespread; as is increasingly being accepted, complex character states are more readily lost than gained. Therefore, we recommend that morphological systematists routinely implement a range of character transformation models to assess the sensitivity of their phylogenetic reconstructions. We discuss the antitropical biogeography of osmeroid fishes in the context of vicariance and transequatorial dispersal.

Animals↗
PubMed2005

Mechanisms of acrylamide induced rodent carcinogenesis.

Acrylamide is a monomer of polyacrylamide, used in biochemistry, in paper manufacture, in water treatment, and as a soil stabilizer. The monomer can cause several toxic effects and has the potential for human exposure either through the environment or from occupational exposure. Recently, additional concern for the potential toxicity of acrylamide in humans has arisen with the finding of acrylamide formation in some processed foods. It has been established that following chronic exposure, rats exhibited an increase in the incidence of adrenal pheochromocytomas, testicular mesotheliomas, thyroid adenomas and mammary neoplasms in F344 rats. This has raised increased concerns regarding the carcinogenic risk to humans from acrylamide exposure. Studies examining the DNA reactivity of acrylamide have been performed and have had differing results. The tissue and organ pattern of neoplastic development seen in the rat following acrylamide exposure is not consistent with that seen with other strictly DNA reactive carcinogens. Based on the pattern of neoplastic development, it appears that acrylamide is targeting endocrine sensitive tissues. In the current monograph, studies on the effect of acrylamide on DNA reactivity and on altered cell growth in the target tissues in the rat are reported. DNA synthesis was examined in F344 rats treated with acrylamide (0, 2, or 15 mg/kg/day) for 7, 14, or 28 days. Acrylamide increased DNA synthesis in the target tissues (thyroid, testicular mesothelium, adrenal medulla) at all doses and time points examined. In contrast, in a non-target tissue (liver), no increase in DNA synthesis was seen. Examination of DNA damage using single cell gel electrophoresis (the Comet assay) showed an increase in DNA damage in the target tissues, but not in non-target tissue (liver). In addition, a cellular transformation model, (the Syrian Hamster Embryo (SHE) cell morphological transformation model), was used to examine potential mechanisms for the observed carcinogenicity of acrylamide. SHE cell studies showed that glutathione (GSH) modulation by acrylamide was important in the cell transformation process. Treatment with a sulfhydryl donor compound (NAC) reduced acrylamide transformation while depletion of GSH (BSO) resulted in an enhancement of transformation. In summary, acrylamide caused both an increase in DNA synthesis and DNA damage in mammalian tissues and cells suggesting that DNA reactivity and cell proliferation, in concert, may contribute to the observed acrylamide-induced carcinogenicity in the rat and has implication on the possible risk for human neoplasm development.

Acrylamide↗
PubMed2000

Modeling the transformation of exogenously supplied cAMP in the perfused rat liver.

A kinetic model describing the behavior of extracellularly supplied cAMP in the perfused rat liver was derived and compared with experimental data. The model was based on the following conditions and assumptions: a) labeled cAMP is being constantly infused (step input); b) permeation of the cell membrane is an essentially irreversible step (k(in) as transfer coefficient); c) the adenine moiety of cAMP incorporates into a nucleotide pool (km1 as transformation coefficient), which cannot permeate the cell membrane; d) the adenine moiety of cAMP can be transferred from the nucleotide pool to a nucleoside + free base pool (km2 as transformation coefficient), which is able to permeate the cell membrane (k(ef) as transfer coefficient). These events were described by a series of differential equations for which an analytical solution was obtained. Total cellular incorporation of label derived from [3H]cAMP was measured in the isolated perfused rat liver. The equations of the model were fitted to these experimental data by means of a least-squares procedure. In the fitting procedure the previously determined k(in) value (0.55 ml min(-1) ml cellular space(-1)) was used. The model is able to describe the experimental data (correlation coefficient = 0.993 +/- 0.008) with km1, km2 and k(ef) values of 17.11, 0.0948 and 1.385 min(-1), respectively. Simulations revealed the following sequence of decreasing intracellular pool sizes: nucleotide pool > nucleoside + free base pool > intracellular cAMP. The intracellular cAMP concentrations correspond to only 3.2% of the extracellular ones. This low proportion explains why it was generally difficult to detect cAMP in the cell space when this compound was added to an isolated cell system. The model and the parameters determined in the present work can be used to predict intracellular cAMP concentrations in the perfused liver for specific extracellular concentrations.

Animals↗
PubMed1981

Yeast transformation: a model system for the study of recombination.

DNA molecules that integrate into yeast chromosomes during yeast transformation do so by homologous recombination. We have studied the way in which circular and linear molecules recombine with homologous chromosomal sequences. We show that DNA ends are highly recombinogenic and interact directly with homologous sequences. Circular hybrid plasmids can integrate by a single reciprocal crossover, but only at a low frequency. Restriction enzyme digestion within a region homologous to yeast chromosomal DNA greatly enhances the efficiency of integration. Furthermore, if two restriction cuts are made within the same homologous sequence, thereby removing an internal segment of DNA, the resulting deleted-linear molecules are still able to transform at a high frequency. Surprisingly, the integration of these gapped-linear molecules results in replacement of the missing segment using chromosomal information. The final structure is identical to that obtained from integration of a circular molecule. The integration of linear and gapped-linear molecules, but not of circular molecules, is blocked by the rad52-1 mutation. Consideration of models for plasmid integration and gene conversion suggests that RAD52 may be involved in the DNA repair synthesis required for these processes. Implications of this work for the isolation of integrative transformants, fine-structure mapping, and the cloning of mutations are discussed.

Base Composition↗
PubMedDate not supplied

Hydrogen bonding in lignin: a Fourier transform infrared model compound study.

Hydrogen bonding plays an important role in the thermal and mechanical properties of biopolymers. To investigate hydrogen bond formation in lignin, an abundant natural polymer found in plants, Fourier transform infrared (FTIR) analysis of various lignin model compounds was performed. Four monomeric model compounds and one dimeric model compound were studied under various conditions. FTIR analysis revealed aliphatic hydroxyl groups form stronger hydrogen bonds than phenolic hydroxyl groups. Further, the dimeric biphenyl-type structure formed significantly stronger intermolecular hydrogen bonds as compared to the other monomeric model compounds. Results from the model compound studies were used to explain the observed complex hydrogen-bonding system present in both softwood and hardwood technical lignins. Together with chemical analysis, we discuss the difference in hydrogen bonding between hardwood and softwood lignin and the observed differences in the glass transition temperature.

Biopolymers↗
PubMed2003

A quantitative model for transforming reflectance spectra into the Munsell color space using cone sensitivity functions and opponent process weights.

This article presents a computational model of the process through which the human visual system transforms reflectance spectra into perceptions of color. Using physical reflectance spectra data and standard human cone sensitivity functions we describe the transformations necessary for predicting the location of colors in the Munsell color space. These transformations include quantitative estimates of the opponent process weights needed to transform cone activations into Munsell color space coordinates. Using these opponent process weights, the Munsell position of specific colors can be predicted from their physical spectra with a mean correlation of 0.989.

Color Perception↗
PubMed2005

Articulated rigid registration for serial lower-limb mouse imaging.

This paper describes a new piecewise rotational transformation model for capturing the articulation of joints such as the hip and the knee. While a simple piecewise rigid model can be applied, such models suffer from discontinuities at the motion boundary leading to both folding and stretching. Our model avoids both of these problems by constructing a provably continuous transformation along the motion interface. We embed this transformation model within the robust point matching framework and demonstrate its successful application to both synthetic data, and to serial x-ray CT mouse images. In the later case, our model captures the articulation of six joints, namely the left/right hip, the left/right knee and the left/right ankle. In the future such a model could be used to initialize non-rigid registrations of images from different subjects, as well as, be embedded in intensity-based and integrated registration algorithms. It could also be applied to human data in cases where articulated motion is an issue (e.g. image guided prostate radiotherapy, lower extremity CT angiography).

Algorithms↗