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 253 records · Page 14Linked to original sources
PubMed2000

Learning cortical topography from spatiotemporal stimuli.

Stimulus representation is a functional interpretation of early sensory cortices. Early sensory cortices are subject to stimulus-induced modifications. Common models for stimulus-induced learning within topographic representations are based on the stimuli's spatial structure and probability distribution. Furthermore, we argue that average temporal stimulus distances reflect the stimuli's relatedness. As topographic representations reflect the stimuli's relatedness, the temporal structure of incoming stimuli is important for the learning in cortical maps. Motivated by recent neurobiological findings, we present an approach of cortical self-organization that additionally takes temporal stimulus aspects into account. The proposed model transforms average interstimulus intervals into representational distances. Thereby, neural topography is related to stimulus dynamics. This offers a new time-based interpretation of cortical maps. Our approach is based on a wave-like spread of cortical activity. Interactions between dynamics and feedforward activations lead to shifts of neural activity. The psychophysical saltation phenomenon may represent an analogue to the shifts proposed here. With regard to cortical plasticity, we offer an explanation for neurobiological findings that other models cannot explain. Moreover, we predict cortical reorganizations under new experimental, spatiotemporal conditions. With regard to psychophysics, we relate the saltation phenomenon to dynamics and interaction in early sensory cortices and predict further effects in the perception of spatiotemporal stimuli.

Animals↗
PubMed2002

Lack of association between CYP2A5 induction and apoptosis in mouse primary hepatocytes.

Upregulation of mouse hepatic cytochrome P450 2A5 (CYP2A5) is a process closely associated with hepatocellular damage and formation of liver tumours. 2-Aminopurine, a protein kinase inhibitor modulating cell cycle control, was recently shown to strongly induce CYP2A5 in mouse hepatocytes. The objective of this study was to determine the association between CYP2A5 induction and apoptosis in mouse primary hepatocytes. Five well-characterised CYP2A5 inducers were tested for their ability to affect apoptosis rate, determined by immunohistochemical in situ 3'-end-labelling technique, in a primary mouse hepatocyte model. Transforming growth factor beta (TGFbeta) was used as a positive (proapoptotic) control. Phenobarbital, pyrazole and the mitogen-activated protein kinase inhibitor PD98059 did not significantly affect apoptosis rate in hepatocytes. Norcocaine induced apoptosis at 6 hr (1.8-fold) and 2-aminopurine 12 hr (1.4-fold) after treatment, which is considerably earlier than peaks in the amount of CYP2A5 mRNA. TGFbeta reduced CYP2A5 marker activity, coumarin 7-hydroxylase by 74%. These results indicate that in a primary hepatocyte model (a) there is no systematic correlation between apoptosis and CYP2A5 induction; (b) phenobarbital does not significantly affect the rate of apoptosis; and (c) the induction of apoptosis caused by the chemicals tested occurs considerable earlier than elevation of CYP2A5 expression. Thus, no causal link appears to exist between induction of CYP2A5 and apoptotic rate.

2-Aminopurine↗
PubMed2004

Robust H(infinity) output feedback control for a class of uncertain Lur'e systems with time-delays.

In this work, the analysis of robust stability and design of robust H infinity output feedback controllers for a class of Lur'e systems with both time-delays and parameter uncertainties were studied. A robust H infinity output feedback controller based on Linear Matrix Inequalities (LMIs) was developed to guarantee the robust stability and H infinity performance of the resultant closed-loop system. The presented design approach is based on the application of descriptor model transformation and Park's inequality for the bounding of cross terms and is expected to be less conservative compared to reported design methods. Finally, illustrative examples are advanced to demonstrate the superiority of the obtained method.

Algorithms↗
PubMed2000

Do the CK18 related proteins change in general in epithelial cancers?

The modulation of cytokeratin 18 during tumor transformation in hepatoma had been previously recognized through a series of biochemical and immunological approaches. Expression of cytokeratin 18 in transitional cell carcinoma comparing with hepatoma was investigated using the hepatoma transformation model. CK18 related molecules were found. In the present study, we design various epithelial cancers with the same model. CK18 related molecules were all evident. Therefore, we suggest that CK18 related proteins would play an important role in tumorigenesis of epithelial cancers.

Carcinoma, Hepatocellular↗
PubMed1991

Altered coupling states between calcium transport and (Ca2+, Mg2+)-ATPase in the AS-30D ascites hepatocarcinoma plasma membrane.

Plasma membrane fractions from normal, regenerating liver and the AS-30D ascites hepatocarcinoma exhibited a high degree of enrichment when a set of plasma membrane enzyme markers were studied in comparison to the ones associated to the mitochondrial and cytosolic compartments. While the (Ca2+, Mg2+)-ATPase observed for the plasma membrane fraction isolated from normal liver showed an activity of 1.2 mumoles/mg/min, the regenerating liver and the AS-30D plasma membrane fractions presented a much lower ATPase activity (0.3 and 0.22 mumoles/mg/min respectively). Despite the differences in ATPase activity observed between models, the plasma membrane fraction from the AS-30D hepatocarcinoma presented a calcium transport activity similar to the value observed for the normal system (5.9 and 5.5 nmoles Ca2+/mg/10 min, respectively). Interestingly, the ATP in equilibrium with Pi exchange experiments carried out with the different plasma membrane fractions revealed that the (Ca2+, Mg2+)-ATPase contained in the plasma membrane from the AS-30D cells shows an exchange activity of 26 nmoles ATP in equilibrium with Pi/mg/min, similar to the one observed fo the enzyme from normal liver (30 nmoles ATP in equilibrium with Pi/mg/min). Our results suggest that the plasma membrane from the transformed model presents a more efficient mechanism to regulate the movement of calcium through the calcium pump, with an optimum expenditure of energy.

Adenosine Triphosphate↗
PubMed2006

The pervasiveness of pharmaceutical expenditure inertia in the OECD countries.

This paper constructs and estimates an economic model for testing statistically the strength of possible 'expenditure inertia' as a plausible reason for rising drug expenditures of the Organization for Economic Cooperation and Development (OECD) countries. The ethical drugs sector in the OECD health care systems is increasingly targeted as the major culprit in the rising cost. Using multiple regression analysis, and the maximum likelihood estimation method, the data of each country (taken from OECD Health Data, 1997) were first tested for functional form optimality with the Box-Cox power family transformations model. Drug expenditure elasticities, at data means, were computed using each country's optimal regression model estimates. The results indicate that the traditionally fitted a priori limited functional form models (e.g., linear, log-log) are not globally consistent with data across countries. The effect of a one-period lagged real per-capita drug expenditure (capturing inertia or habit persistence) on current period real per-capita prescription expenditure is statistically significant in most countries. Pharmaceutical demands are inelastic, and tend to behave like a necessity, as expected. Since the significant effects of economic, demographic, and other drivers of high drug spending differ across countries, country-specific implications and policy suggestions for cost controls ought to differ.

Aged↗
PubMed2002

Influence of anthropogenic aerosol on cloud optical depth and albedo shown by satellite measurements and chemical transport modeling.

The Twomey effect of enhanced cloud droplet concentration, optical depth, and albedo caused by anthropogenic aerosols is thought to contribute substantially to radiative forcing of climate change over the industrial period. However, present model-based estimates of this indirect forcing are highly uncertain. Satellite-based measurements would provide global or near-global coverage of this effect, but previous efforts to identify and quantify enhancement of cloud albedo caused by anthropogenic aerosols in satellite observations have been limited, largely because of strong dependence of albedo on cloud liquid water path (LWP), which is inherently highly variable. Here we examine satellite-derived cloud radiative properties over two 1-week episodes for which a chemical transport and transformation model indicates substantial influx of sulfate aerosol from industrial regions of Europe or North America to remote areas of the North Atlantic. Despite absence of discernible dependence of optical depth or albedo on modeled sulfate loading, examination of the dependence of these quantities on LWP readily permits detection and quantification of increases correlated with sulfate loading, which are otherwise masked by variability of LWP, demonstrating brightening of clouds because of the Twomey effect on a synoptic scale. Median cloud-top spherical albedo was enhanced over these episodes, relative to the unperturbed base case for the same LWP distribution, by 0.02 to 0.15.

Aerosols↗
PubMed2026

Harnessing the Power of Large Language Models for Drug Discovery: A Systematic Review of Current Applications and Future Directions.

INTRODUCTION: The demand for inventive approaches to drug discovery has increased due to the rising costs, time, and failure rates in pharmaceutical research. Large Language Models (LLMs), with their sophisticated natural language processing and generative capabilities, have become potent instruments that have the potential to revolutionize biomedical research. The function of LLMs in different phases of drug development is methodically examined in this article. METHODS: The PRISMA 2020 principles were adhered to in this systematic study. A thorough search for research published between 2018 and 2025 was done using PubMed, Scopus, Web of Science, and Google Scholar. The search terms "large language model," "transformer," "drug discovery," and important sub-domains (such as "de-novo design" and "ADMET") were merged, and two reviewers independently screened the results. Predetermined inclusion and exclusion criteria were used to filter studies for relevance. 98 studies out of the 1,285 records that were initially retrieved met the requirements for the final qualitative synthesis. RESULTS: 98 studies that demonstrated the use of LLMs in various drug discovery domains were found during the review. These covered molecular generation, genomics, protein-ligand modeling, ADME/T and toxicity profiling, drug-target interaction and DTI prediction, and biomedical text mining. 42 different LLM-based tools were mapped, including BioBERT, SciSpacy, Drug- LLM, DNA-BERT, GPT-4, and ChatGPT. Predictive accuracy, hypothesis creation, target prioritization, and multi-modal data integration all showed notable gains with these techniques. DISCUSSION: By providing scalable, precise, and effective solutions for data-driven drug discovery, LLMs are revolutionizing the pharmaceutical industry. They allow for the creation of hypotheses and individualized insights across multi-modal biological data, and they perform better than conventional approaches in a number of subdomains. Improvements in performance were task-dependent; the most consistent gains occurred for biomedical text mining, disease-genedrug relationship mapping and drug-target interaction prediction tasks. Yet most evidence for clinical applications is still derived from retrospective studies and benchmark datasets, suggesting a higher need for prospective validation. CONCLUSION: There is revolutionary potential in incorporating LLMs into drug discovery processes. Clinical translation and regulatory uptake will depend heavily on collaborative validation, ethical deployment, and standardization as models become more multimodal and interpretable. Before normal use, extensive prospective benchmarking and head-to-head comparisons with established chemoinformatics pipelines are necessary.

De novo design↗
PubMed2004

Adsorption isotherm models for basic dye adsorption by peat in single and binary component systems.

Colored effluents from textile industries are a problem in many rivers and waterways. Prediction of dye adsorption capacities is important in design considerations. The sorption of three basic dyes, namely Basic blue 3, Basic yellow 21, and Basic red 22, onto peat is reported. Equilibrium sorption isotherms have been measured for the three single-component systems. Equilibrium was achieved after 21 days. The experimental isotherm data were analyzed using Langmuir, Freundlich, Redlich-Peterson, Tempkin, and Toth isotherm equations. A detailed error analysis has been undertaken to investigate the effect of using different error criteria for the determination of the single-component isotherm parameters and hence obtain the best isotherm and isotherm parameters which describe the adsorption process. The linear transform model provided the highest R(2) regression coefficient with the Redlich-Peterson model. The Redlich-Peterson model also yielded the best fit to experimental data for all three dyes using the nonlinear error functions. An extended Langmuir model has been used to predict the isotherm data for the binary systems using the single component data. The correlation between theoretical and experimental data had only limited success due to competitive and interactive effects between the dyes and the dye-surface interactions.

Journal Article↗
PubMed2004

Insufficient p65 phosphorylation at S536 specifically contributes to the lack of NF-kappaB activation and transformation in resistant JB6 cells.

NF-kappaB activation is required for TNF-alpha-induced transformation of JB6 mouse epidermal cells. Deficient activation of p65 contributes to the lack of NF-kappaB activation in transformation-resistant (P-) cells. We hypothesized that the differential NF-kappaB activation involves differential p65 phosphorylation arising from enzyme activity differences. Here we show that TNF-alpha induces greater ERK-dependent p65 phosphorylation at S536 in transformation sensitive (P+) cells than in P- cells. Our results establish that limited ERK content contributes to a low IkappaB kinase (IKKbeta) level, in turn resulting in insufficient p65 phosphorylation at S536 upon TNF-alpha stimulation in P- cells. Phosphorylation of p65 at S536 appears to play a role in TNF-alpha-induced p65 DNA binding and recruitment of p300 to the p65 complex as well as in release of p65 bound to HDAC1 and 3. Blocking p65 phosphorylation at S536, but not at S276 or S529, abolishes p65 transactivational activity. Over-expression of p65 but not p65 phosphorylation mutant (S536A) in transformation-resistant P- cells renders these cells sensitive to TNF-alpha-induced transformation. Over-expression of p65 phosphorylation mimics p65-S536D or p65-S536E in P- cells and also rescues the transformation response. These findings provide direct evidence that phosphorylation of p65 at S536 is required for TNF-alpha-induced NF-kappaB activation in the JB6 transformation model. The lack of NF-kappaB activation seen in P- cells can be attributed to an insufficient level of p65 phosphorylation on S536 that arises from insufficient IKKbeta that in turn arises from insufficient ERK. Thus, p65 phosphorylation at S536 offers a potential molecular target for cancer prevention.

Active Transport, Cell Nucleus↗
PubMed1999

A three-dimensional kinematic model of the human long finger and the muscles that actuate it.

A kinematic model of the human long finger and the six muscles that actuate it is presented. The model transforms finger pose into estimates of muscle excursions and fingertip location. The effects of abduction/adduction about the metacarpo-phalangeal joint are accounted for, as are the effects of flexion of the three finger joints. A set of parameters are provided which approximate kinematics of the segments and muscles of a cadaver finger over the range of finger poses humans normally assume.

Biomechanical Phenomena↗
PubMed1998

Coordinated force production in multi-finger tasks: finger interaction and neural network modeling.

During maximal voluntary contraction (MVC) with several fingers, the following three phenomena are observed: (1) the total force produced by all the involved fingers is shared among the fingers in a specific manner (sharing); (2) the force produced by a given finger in a multi-finger task is smaller than the force generated by this finger in a single-finger task (force deficit); (3) the fingers that are not required to produce any force by instruction are involuntary activated (enslaving). We studied involuntary force production by individual fingers (enslaving effects, EE) during tasks when (an)other finger(s) of the hand generated maximal voluntary pressing force in isometric conditions. The subjects (n = 10) were instructed to press as hard as possible on the force sensors with one, two, three and four fingers acting in parallel in all possible combinations. The EE were (A) large, the slave fingers always producing a force ranging from 10.9% to 54.7% of the maximal force produced by the finger in the single-finger task; (B) nearly symmetrical; (C) larger for the neighboring fingers; and (D) non-additive. In most cases, the EE from two or three fingers were smaller than the EE from at least one finger (this phenomenon was coined occlusion). The occlusion cannot be explained only by anatomical musculo-tendinous connections. Therefore, neural factors contribute substantially to the EE. A neural network model that accounts for all the three effects has been developed. The model consists of three layers: the input layer that models a central neural drive; the hidden layer modeling transformation of the central drive into an input signal to the muscles serving several fingers simultaneously (multi-digit muscles); and the output layer representing finger force output. The output of the hidden layer is set inversely proportional to the number of fingers involved. In addition, direct connections between the input and output layers represent signals to the hand muscles serving individual fingers (uni-digit muscles). The network was validated using three different training sets. Single digit muscles contributed from 25% to 50% of the total finger force. The master matrix and the enslaving matrix were computed; they characterize the ability of a given finger to enslave other fingers and its ability to be enslaved. Overall, the neural network modeling suggests that no direct correspondence exists between neural command to an individual finger and finger force. To produce a desired finger force, a command sent to an intended finger should be scaled in accordance with the commands sent to the other fingers.

Adult↗
PubMed1994

Classical analytical methods for detecting matching effects on treatment outcome.

This article presents a classical approach for analyzing repeated measures designs with specific application to treatment matching studies. The generic treatment matching hypothesis is formulated under the multivariate general linear model, transforming the dependent variables to account for the repeated measures structure of the data. Issues of primary importance in the use of this approach (such as correcting for inflated Type I error and robustness of statistical tests to parametric assumptions) are discussed. The article concludes with an assessment of the strengths and weaknesses of this approach compared with alternative approaches.

Alcoholism↗
PubMed1996

Neoplastic transformation and DNA-binding of 4,4'-methylenebis(2-chloroaniline) in SV40-immortalized human uroepithelial cell lines.

The tumorigenic transformation of certain occupationally significant chemicals, such as N-hydroxy-4-4'-methylenebis[2-chloroaniline] (N-OH-MOCA), N-hydroxy-ortho-toluidine (N-OH-OT), 2-phenyl-1,4-benzoquinone (PBQ) and N-hydroxy-4-aminobiphenyl (N-OH-ABP) were tested in vitro using the well established SV40-immortalized human uroepithelial cell line SV-HUC.PC. SV-HUC cells were exposed in vitro to varying concentrations of N-OH-MOCA, N-OH-OT, N-OH-ABP and PBQ that caused approximately 25% and 75% cytotoxicity. The carcinogen treated cells were propagated in culture for about six weeks and subsequently injected subcutaneously into athymic nude mice. Two of the fourteen different groups of SV-HUC.PC treated with different concentrations of N-OH-MOCA, and one of the three groups exposed to N-OH-ABP, formed carcinomas in athymic nude mice. 32P-postlabeling analyses of DNA isolated from SV-HUC.PC after exposure to N-OH-MOCA revealed one major and one minor adduct. The major adduct has been identified as the N-(deoxyadenosin-3',5'-bisphospho-8-yl)-4-amino-3-chlorob enz yl alcohol (pdAp-ACBA) and the minor adduct as N-(deoxyadenosin-3',5'-bisphospho-8-yl)-4-amino-3-chlorot oluene (pdApACT). Furthermore, SV-HUC.PC cytosols catalyzed the binding of N-OH-MOCA to DNA, in the presence of acetyl-CoA, to yield similar adducts. The same adducts were also formed by chemical interaction of N-OH-MOCA with calf thymus DNA, suggesting that the aryl nitrenium ion may be the ultimate reactive species responsible for DNA binding. The tumorigenic activity of N-OH-MOCA in this highly relevant in vitro transformation model, coupled with the findings that SV-HUC.PC cells formed DNA-adducts in vitro and contained enzyme systems that activated N-OH-MOCA to reactive electrophilic species that bound to DNA, strongly suggest that MOCA could be a human bladder carcinogen. These findings are consistent with the International Agency for Research on Cancer's classification of MOCA as a probable human carcinogen.

Animals↗
PubMed2004

Down-regulation of Betaig-h3 gene is involved in the tumorigenesis in human bronchial epithelial cells induced by heavy-ion radiation.

High-energy (HZE) heavy ions, when compared to low-LET radiation, are highly effective in inducing gene mutation, chromosomal aberrations and neoplastic transformation. However, the underlying molecular mechanisms are not clearly understood. We have recently shown that the down-regulation of Betaig-h3 expression is causally linked to the tumorigenic phenotype of papillomavirus-immortalized human bronchial epithelial (BEP2D) cells treated with high-LET alpha-particle radiation. Using the BEP2D cell culture system, a radiation-induced transformation model has been established by a single 60-cGy dose of (56)Fe heavy-ion radiation. To determine whether the Betaig-h3 gene is involved in (56)Fe ion-induced tumorigenesis, the expression levels of the Betaig-h3 gene in tumorigenic cell lines and the ability of in vivo tumor suppression through the reintroduction of the Betaig-h3 gene in tumorigenic cells were determined. We found that the expression level of this gene is markedly decreased in three tumorigenic cell lines ((56)FeT1-T3) compared with parental BEP2D cells. Ectopic expression of its cDNA in the (56)FeT2 tumorigenic cells significantly suppressed their tumorigenicity. Although biologically active TGFB1 is elevated in two of three tumorigenic cell lines, all these cell lines are resistant to the induction of Betaig-h3 expression by incubating the transformed cells with exogenous TGFB1 relative to control cells. Our data strongly suggest that down-regulation of Betaig-h3 expression results from the defect in the TGFB1 signaling pathway and plays a pivotal role in the tumorigenic process induced by (56)Fe heavy-ion radiation.

Bronchi↗
PubMedDate not supplied

Designing a mixed methods study in primary care.

BACKGROUND: Mixed methods or multimethod research holds potential for rigorous, methodologically sound investigations in primary care. The objective of this study was to use criteria from the literature to evaluate 5 mixed methods studies in primary care and to advance 3 models useful for designing such investigations. METHODS: We first identified criteria from the social and behavioral sciences to analyze mixed methods studies in primary care research. We then used the criteria to evaluate 5 mixed methods investigations published in primary care research journals. RESULTS: Of the 5 studies analyzed, 3 included a rationale for mixing based on the need to develop a quantitative instrument from qualitative data or to converge information to best understand the research topic. Quantitative data collection involved structured interviews, observational checklists, and chart audits that were analyzed using descriptive and inferential statistical procedures. Qualitative data consisted of semistructured interviews and field observations that were analyzed using coding to develop themes and categories. The studies showed diverse forms of priority: equal priority, qualitative priority, and quantitative priority. Data collection involved quantitative and qualitative data gathered both concurrently and sequentially. The integration of the quantitative and qualitative data in these studies occurred between data analysis from one phase and data collection from a subsequent phase, while analyzing the data, and when reporting the results. DISCUSSION: We recommend instrument-building, triangulation, and data transformation models for mixed methods designs as useful frameworks to add rigor to investigations in primary care. We also discuss the limitations of our study and the need for future research.

Health Services Research↗
PubMed1985

beta-Carotene reduces sister chromatid exchanges induced by chemical carcinogens in mouse mammary cells in organ culture.

Present studies in the mammary epithelial cell transformation model in organ culture showed that presence of beta-carotene during the 24 hr treatment (initiation stage) of the glands with the carcinogens, 7,12-dimethylbenz[a]anthracene (DMBA), N-nitrosodiethylamine (DENA) and N-methylnitrosourea (MNU), caused a highly significant (P less than 0.001-0.01) reduction of SCE induced by the same carcinogens. In contrast, 4-hydroxyphenyl retinamide (4-HPR) which is known to act at the promotional stage of carcinogenesis did not show any significant reduction of SCE. Thus findings suggest that beta-carotene can modify the DNA damaging effect of the carcinogens and thereby may also prevent the initiation of the carcinogenic process.

9,10-Dimethyl-1,2-benzanthracene↗
PubMed2004

ARTSTREAM: a neural network model of auditory scene analysis and source segregation.

Multiple sound sources often contain harmonics that overlap and may be degraded by environmental noise. The auditory system is capable of teasing apart these sources into distinct mental objects, or streams. Such an 'auditory scene analysis' enables the brain to solve the cocktail party problem. A neural network model of auditory scene analysis, called the ARTSTREAM model, is presented to propose how the brain accomplishes this feat. The model clarifies how the frequency components that correspond to a given acoustic source may be coherently grouped together into a distinct stream based on pitch and spatial location cues. The model also clarifies how multiple streams may be distinguished and separated by the brain. Streams are formed as spectral-pitch resonances that emerge through feedback interactions between frequency-specific spectral representations of a sound source and its pitch. First, the model transforms a sound into a spatial pattern of frequency-specific activation across a spectral stream layer. The sound has multiple parallel representations at this layer. A sound's spectral representation activates a bottom-up filter that is sensitive to the harmonics of the sound's pitch. This filter activates a pitch category which, in turn, activates a top-down expectation that is also sensitive to the harmonics of the pitch. Resonance develops when the spectral and pitch representations mutually reinforce one another. Resonance provides the coherence that allows one voice or instrument to be tracked through a noisy multiple source environment. Spectral components are suppressed if they do not match harmonics of the top-down expectation that is read-out by the selected pitch, thereby allowing another stream to capture these components, as in the 'old-plus-new heuristic' of Bregman. Multiple simultaneously occurring spectral-pitch resonances can hereby emerge. These resonance and matching mechanisms are specialized versions of Adaptive Resonance Theory, or ART, which clarifies how pitch representations can self-organize during learning of harmonic bottom-up filters and top-down expectations. The model also clarifies how spatial location cues can help to disambiguate two sources with similar spectral cues. Data are simulated from psychophysical grouping experiments, such as how a tone sweeping upwards in frequency creates a bounce percept by grouping with a downward sweeping tone due to proximity in frequency, even if noise replaces the tones at their intersection point. Illusory auditory percepts are also simulated, such as the auditory continuity illusion of a tone continuing through a noise burst even if the tone is not present during the noise, and the scale illusion of Deutsch whereby downward and upward scales presented alternately to the two ears are regrouped based on frequency proximity, leading to a bounce percept. Since related sorts of resonances have been used to quantitatively simulate psychophysical data about speech perception, the model strengthens the hypothesis that ART-like mechanisms are used at multiple levels of the auditory system. Proposals for developing the model to explain more complex streaming data are also provided.

Acoustic Stimulation↗