Search PubMed⌕ Search

SEARCH · Search PubMed

Results for “ARCHITECTURE”

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 523 records · Page 29Linked to original sources
PubMed2003

A hierarchical method for finding optimal architecture and weights using evolutionary least square based learning.

In this paper, we present a novel approach of implementing a combination methodology to find appropriate neural network architecture and weights using an evolutionary least square based algorithm (GALS).1 This paper focuses on aspects such as the heuristics of updating weights using an evolutionary least square based algorithm, finding the number of hidden neurons for a two layer feed forward neural network, the stopping criterion for the algorithm and finally some comparisons of the results with other existing methods for searching optimal or near optimal solution in the multidimensional complex search space comprising the architecture and the weight variables. We explain how the weight updating algorithm using evolutionary least square based approach can be combined with the growing architecture model to find the optimum number of hidden neurons. We also discuss the issues of finding a probabilistic solution space as a starting point for the least square method and address the problems involving fitness breaking. We apply the proposed approach to XOR problem, 10 bit odd parity problem and many real-world benchmark data sets such as handwriting data set from CEDAR, breast cancer and heart disease data sets from UCI ML repository. The comparative results based on classification accuracy and the time complexity are discussed.

Algorithms↗
PubMed1999

A modular neural network architecture for pattern classification based on different feature sets.

We propose a novel connectionist method for the use of different feature sets in pattern classification. Unlike traditional methods, e.g., combination of multiple classifiers and use of a composite feature set, our method copes with the problem based on an idea of soft competition on different feature sets developed in our earlier work. An alternative modular neural network architecture is proposed to provide a more effective implementation of soft competition on different feature sets. The proposed architecture is interpreted as a generalized finite mixture model and, therefore, parameter estimation is treated as a maximum likelihood problem. An EM algorithm is derived for parameter estimation and, moreover, a model selection method is proposed to fit the proposed architecture to a specific problem. Comparative results are presented for the real world problem of speaker identification.

Humans↗
PubMed2002

Contribution of laminar myofiber architecture to load-dependent changes in mechanics of LV myocardium.

The ventricular myocardium consists of a syncytium of myocytes organized into branching, transmurally oriented laminar sheets approximately four cells thick. When systolic deformation is expressed in an axis system determined by the anatomy of the laminar architecture, laminar sheets of myocytes shear and laterally extend in an approximately radial direction. These deformations account for ~90% of normal systolic wall thickening in the left ventricular free wall. In the present study, we investigated whether the changes in systolic and diastolic function of the sheets were sensitive to alterations in systolic and diastolic load. Our results indicate that there is substantial reorientation of the laminar architecture during systole and diastole. Moreover, this reorientation is both site and load dependent. Thus as end-diastolic pressure is increased and the left ventricular wall thins, sheets shorten and rotate away from the radial direction due to transverse shearing, opposite of what occurs in systole. Both mechanisms of thickening contribute substantially to normal left ventricular wall function. Whereas the relative contributions of shear and extension are comparable at the base, sheet shear is the predominant factor at the apex. The magnitude of shortening/extension and shear increases with preload and decreases with afterload. These findings underscore the essential contribution of the laminar myocardial architecture for normal ventricular function throughout the cardiac cycle.

Animals↗
PubMed1994

Mechanical equilibrium determines the fractal fiber architecture of aortic heart valve leaflets.

In this work, the structure of the aortic valve is derived from its function, which (in the closed-valve configuration) is to support a uniform pressure load. It is assumed that this load is transferred to the aortic wall by a one-parameter family of fibers under tension. The equation of equilibrium for this fiber structure turns out to be equivalent to the equation of motion of vortex lines in the self-induction approximation. The method of Buttke (J. Comput. Phys. 76:301-326, 1988) is used to solve these equations and, hence, to determine the fiber architecture of the aortic leaflets. Because of a singularity at the center of the aortic valve, the computed fiber architecture has a fractal character with increasing complexity at progressively smaller scales. The computed fiber architecture resembles the branching braided structure of the collagen fibers that support the real aortic valve.

Aortic Valve↗
PubMed1998

Muscle fiber architecture of the dog diaphragm.

Previous measurements of muscle thickness and length ratio of costal diaphragm insertions in the dog (A. M. Boriek and J. R. Rodarte. J. Appl. Physiol. 77: 2065-2070, 1994) suggested, but did not prove, discontinuous muscle fiber architecture. We examined diaphragmatic muscle fiber architecture using morphological and histochemical methods. In 15 mongrel dogs, transverse sections along the length of the muscle fibers were analyzed morphometrically at x20, by using the BioQuant System IV software. We measured fiber diameters, cross-sectional fiber shapes, and cross-sectional area distributions of fibers. We also determined numbers of muscle fibers per cross-sectional area and ratio of connective tissue to muscle fibers along a course of the muscle from near the chest wall (CW) to near the central tendon (CT) for midcostal left and right hemidiaphragms, as well as ventral, middle, and dorsal regions of the left costal hemidiaphragm. In six other mongrel dogs, the macroscopic distribution of neuromuscular junctions (NMJ) on thoracic and abdominal diaphragm surfaces was determined by staining the intact diaphragmatic muscle for acetylcholinesterase activity. The average major diameter of muscle fibers was significantly smaller, and the number of fibers was significantly larger midspan between CT and CW than near the insertions. The ratio of connective tissues to muscle fibers was largest at CW compared with other regions along the length of the muscle. The diaphragm is transversely crossed by multiple scattered NMJ bands with fairly regular intervals offset in adjacent strips. Muscle fascicles traverse two to five NMJ, consistent with fibers that do not span the entire fascicle from CT to CW. These results suggest that the diaphragm has a discontinuous fiber architecture in which contractile forces may be transmitted among the muscle fibers through the connective tissue adjacent to the fibers.

Animals↗
PubMed2006

Early skeletal muscle hypertrophy and architectural changes in response to high-intensity resistance training.

The onset of whole muscle hypertrophy in response to overloading is poorly documented. The purpose of this study was to assess the early changes in muscle size and architecture during a 35-day high-intensity resistance training (RT) program. Seven young healthy volunteers performed bilateral leg extension three times per week on a gravity-independent flywheel ergometer. Cross-sectional area (CSA) in the central (C) and distal (D) regions of the quadriceps femoris (QF), muscle architecture, maximal voluntary contraction (MVC), and electromyographic (EMG) activity were measured before and after 10, 20, and 35 days of RT. By the end of the training period, MVC and EMG activity increased by 38.9 +/- 5.7 and 34.8% +/- 4.7%, respectively. Significant increase in QF CSA (3.5 and 5.2% in the C and D regions, respectively) was observed after 20 days of training, along with a 2.4 +/- 0.7% increase in fascicle length from the 10th day of training. By the end of the 35-day training period, the total increase in QF CSA for regions C and D was 6.5 +/- 1.1 and 7.4 +/- 0.8%, respectively, and fascicle length and pennation angle increased by 9.9 +/- 1.2 and 7.7 +/- 1.3%, respectively. The results show for the first time that changes in muscle size are detectable after only 3 wk of RT and that remodeling of muscle architecture precedes gains in muscle CSA. Muscle hypertrophy seems to contribute to strength gains earlier than previously reported; flywheel training seems particularly effective for inducing these early structural adaptations.

Adolescent↗
PubMed1988

Fiber architecture and histochemistry in the cat neck muscle, biventer cervicis.

1. Biventer cervicis (BC) is an anatomically complex muscle that is divided by tendinous inscriptions into five in-series compartments of motor units. We have analyzed the fiber architecture and fiber-type composition of these different compartments using microdissection and histochemical methods. 2. BC narrows as it runs rostrally, but its in-series compartments have similar cross-sectional areas. The tapered shape of BC comes about because tendinous inscriptions and the tendon of insertion are oriented obliquely and muscle fibers attach in a progressively offset fashion from the medial to the lateral muscle edge. 3. Individual compartments of BC differ from one another in their architecture. The rostral two compartments (1 and 2) contain fibers of similar length that run between two plates of tendinous tissue. Compartments 3 and 4 are divided into two or three in-parallel subvolumes whose fiber bundles differ in their lengths and sites of attachment. Compartment 5 is the most variable in its structure. In some cats it is separated from compartment 4 by a tendinous inscription, but in other cats, it blends with a dorsomedial part of compartment 4 to form a single subvolume. 4. The relative lengths of fibers in different compartments were analyzed when the head and neck were held in different postures. Fibers in rostromedial regions were stretched more effectively when the head was flexed at suboccipital joints, and appeared to be less sensitive to movements at lower cervical joints. Movements across lower cervical joints produced substantial length changes in caudolateral parts of BC. 5. Muscle fibers of different histochemical types were not distributed evenly within each muscle compartment. Slow, oxidative (SO) fibers accounted for the majority of fibers near the nuchal midline but for only 30%-45% of fibers in lateral muscle regions. Proportions of fast, glycolytic (FG) fibers were greatest in lateral regions. Fast, oxidative-glycolytic (FOG) fibers were distributed quite uniformly throughout each compartment. 6. The specialized architecture of BC may shape its physiological capabilities. The complex internal structures of different compartments may alter the length-tension properties of BC.(ABSTRACT TRUNCATED AT 400 WORDS)

Adenosine Triphosphatases↗
PubMed2026

Mechanistic Perspectives From Genomics and Pangenomics of Medicinal and Aromatic Plants: Linking Genome Architecture to Phytochemical Diversity.

Medicinal and aromatic plants (MAPs) produce a remarkable diversity of specialized metabolites with significant pharmaceutical, nutraceutical, and industrial value. Although advances in long-read sequencing, chromosome-scale genome assembly, and pangenomics have greatly expanded genomic resources, the mechanistic links between genome architecture and phytochemical diversity remain incompletely understood. The present review synthesizes current evidence describing how structural genomic variation may contribute to phytochemical diversity, while acknowledging that many proposed genome-to-metabolite relationships require further experimental validation. Examples illustrate how genome architecture is associated with specialized-metabolite biosynthesis through multiple regulatory processes. However, the strength of supporting evidence varies considerably among MAP species. Moreover, relatively few genome-to-metabolite relationships have been confirmed through direct functional validation. We further discuss how pangenomics, multiomics integration, genome editing, synthetic biology, and artificial intelligence support the discovery, validation, and engineering of specialized metabolic pathways. Casual conclusions are evaluated according to the strength of available evidence, highlighting where causal relationships have been experimentally established and where conclusions remain primarily association-based. Overall, this review provides an integrated conceptual and evidence-based perspective summarizing proposed relationships between genome architecture and phytochemical diversity and outlines future priorities for functional genomics, precision breeding, metabolic engineering, and sustainable utilization of MAPs.

artificial intelligence↗
PubMed1997

Skeletal muscle architecture and fiber-type distribution with the multiple bellies of the mouse extensor digitorum longus muscle.

The purpose of this study was to describe the extent to which architectural and fiber-type characteristics of the four bellies of the mouse extensor digitorum longus (EDL) suggest specialization of the digits, and to mathematically model the functional effects of the structural properties. Six mice were perfused in situ with glutaraldehyde while the lower limb was positioned approximately in the neutral position. After perfusion, lower limbs were removed and placed in glutaraldehyde until the EDL was dissected from the limb and separated into individual muscle bellies corresponding to each digit for architectural determination. The results showed that the muscle belly of digit 5 tended to be different from the muscle bellies of digits 2-4 for many architectural characteristics. Muscle mass, physiological cross-sectional area, muscle length, and fiber length were all significantly greater in digit 5. Proximal tendon length was also significantly longer in digit 5, and distal tendon length, as well as total tendon length, were significantly shorter in digit 5. Sarcomere length was shortest at the proximal end of the muscle and longest, 60-80%, toward the distal end. Fiber type distribution was about 60% FOG, 39% FG with only 1% SO fibers in all muscle bellies. Muscle-tendon modeling illustrated that peak force and maximal shortening velocity were greatest in digit 5. Inclusion of the tendon in the model resulted in a 10% shift of the force-length curve to longer lengths. Assuming muscle structure is matched to function, we speculate that digit 5 of the mouse EDL bears higher loads over a greater excursion during locomotion compared to the remaining digits.

Anatomy, Artistic↗
PubMed1985

Vascular architecture modifications in the steroid-induced polycystic kidney.

This paper presents a study of the renal vascular architecture in polycystic kidneys induced in newborn rabbits by a single injection of methylprednisolone acetate. Ink perfusion and polyester resin corrosion casts were utilized, and the vasculature was studied by light and scanning electron microscopy. Vascular alterations were observed in the two main evolutive stages of the renal polycystosis: the stage of tubular cysts and the stage of glomerular cysts. These alterations were compared with the normal vascular pattern of control kidneys. At the tubular cyst stage the sinusoidal vascular cortical plexus persisted for a long period of time and the superficial cortical glomerular tufts were poorly developed. At the glomerular cyst stage, the vascular architecture of the superficial cortex showed numerous alterations. The glomerular tufts were small and their structural pattern was affected. Frequently the hilum was very wide and the afferent and efferent arterioles were on opposite sides of the glomerular tufts. The afferent arterioles were variable in size and length. Sometimes the glomeruli showed several afferent arterioles. There were numerous cases of double efferent arterioles in the superficial glomeruli of the polycystic kidneys. Various types of aglomerular shunts were observed regardless of the cystic stage or age. These data indicate that, in addition to the cysts, the corticoid-induced polycystic kidneys showed alterations of the renal vascular architecture which are interpreted as a malformation. We suggest, on the basis of these morphologic observations, that the pathogenesis of the polycystic kidneys is not related to the vascular alterations.

Animals↗
PubMedDate not supplied

Cytology of grade 1 papillary transitional cell carcinoma. A comparison of cytologic, architectural and morphometric criteria in cystoscopically obtained urine.

OBJECTIVE: To assess the diagnostic criteria for grade 1 papillary transitional cell carcinoma (TCC) in cystoscopically obtained urine. STUDY DESIGN: We compared the sensitivity, specificity and positive predictive value of cytologic, architectural and morphometric (primarily architectural) criteria in 177 specimens with corresponding biopsy follow-up. RESULTS: Sensitivities ranged from 22% to 44%, specificities from 69% to 85% and positive predictive values from 59% to 66%. CONCLUSION: The currently described cytologic, architectural and morphometric criteria are inadequate for the identification of grade 1 papillary TCC, and the cytologic diagnosis of grade 1 papillary TCC in cystoscopically obtained urine remains unreliable.

Biopsy↗
PubMed2000

Altered myocardial microvascular 3D architecture in experimental hypercholesterolemia.

BACKGROUND: Experimental hypercholesterolemia (HC) impairs intramyocardial microvascular function. However, whether this is associated with alterations in microvascular architecture remained unknown. Using a novel 3D micro-CT scanner, we tested the hypothesis that HC is associated with an alteration in the microvascular architecture. METHODS AND RESULTS: Pigs were euthanized after 12 weeks of either normal (n=6) or 2% HC (n=6) diet. The hearts were excised and the coronary arteries injected with a radiopaque contrast material. Myocardial samples were scanned with micro-CT, and 3D images were reconstructed with 21-microm cubic voxels. The myocardium was tomographically subdivided into subepicardium and subendocardium, and microvessels (<500 microm in diameter) were counted in situ within each region. In the subendocardium of HC pigs, the intramyocardial density of microvessels was significantly higher than in normal animals (1221.4+/-199.7 versus 758.3+/-90.8 vessels/cm(3), P:<0.05) because of an increase in the number of microvessels <200 microm in diameter (1214.4+/-199.7 versus 746. 6+/-101.5 vessels/cm(3), P:<0.05). The subepicardial vascular density was similar in both groups. CONCLUSIONS: -HC has differential effects on the spatial density of the subendocardial microvasculature that may play a role in regulation and/or spatial distribution of myocardial blood flow. This study also demonstrates the feasibility of studying myocardial microvascular architecture with micro-CT in pathophysiological states.

Animals↗
PubMed1996

Hybrid modeling, HMM/NN architectures, and protein applications.

We describe a hybrid modeling approach where the parameters of a mode are calculated and modulated by another model, typically a neural network (NN), to avoid both overfitting and underfitting. We develop the approach for the case of Hidden Markov Models (HMMs), by deriving a class of hybrid HMM/NN architectures. These architectures can be trained with unified algorithms that blend HMM dynamic programming with NN backpropagation. In the case of complex data, mixtures of HMMs or modulated HMMs must be used. NNs can then be applied both to the parameters of each single HMM, and to the switching or modulatation of the models, as a function of input or context. Hybrid HMM/NN architectures provide a flexible NN parameterization for the control of model structure and complexity. At the same time, they can capture distributions that, in practice, are inaccessible to single HMMs. The HMM/NN hybrid approach is tested, in its simplest form, by constructing a model of the immunoglobulin protein family. A hybrid model is trained, and a multiple alignment derived, with less than a fourth of the number of parameters used with previous single HMMs.

Algorithms↗
PubMed1997

Effects of esophageal pressure monitoring on sleep architecture.

Esophageal pressure (Pes) monitoring can be performed during polysomnography with a thin, water-filled catheter connected to a transducer. The resulting quantitative assessment of respiratory effort aids in the diagnosis of sleep-related breathing disorders, but little is known about the potential effect of the procedure on sleep architecture. We monitored Pes during polysomnography in 155 patients and compared their sleep architecture with that of 155 matched patients who slept without esophageal catheters. Cases were matched for age, sex, and severity of respiratory disturbance during sleep. Esophageal manometry was associated with small but statistically significant (p < 0.05) decrements in total recording time, total sleep time, sleep efficiency, percent Stage 2 sleep, and percent rapid-eye-movement (REM) sleep, and with increases in latency to REM sleep, latency to persistent sleep, and percent Stage 3/4 sleep. The differences were of such small magnitude that their clinical significance is doubtful. The number of awakenings per hour of sleep, latency to sleep onset, and percent Stage 1 sleep were no different when esophageal manometry was used. We conclude that the effects of monitoring Pes on sleep architecture are minimal, and that the decision of whether or not to use the technique can be based to a large extent on whether quantitative information about respiratory effort will be useful.

Adolescent↗
PubMed2000

Sleep architecture and respiratory disturbances in children with obstructive sleep apnea.

Little is known regarding sleep architecture in children with the obstructive sleep apnea syndrome (OSAS). We hypothesized that sleep architecture was normal, and that apnea increased over the course of the night, in children with OSAS. We analyzed polysomnographic studies from 20 children with OSAS and 10 control subjects. Sleep architecture was similar between the groups. Of obstructive apneas 55% occurred during rapid eye movement (REM) sleep. The apnea index, apnea duration, and degree of desaturation were greater during REM than non-REM sleep. OSAS data from the first and third periods of the night (periods A and C) were compared. Both the overall and the REM apnea index increased between periods A and C (11 to 25/h, p < 0.02; and 24 to 51/h, p < 0.01, respectively). There was no difference in Sa(O(2)) over time. Spontaneous arousals, but not respiratory-related arousals, were more frequent during non-REM than REM sleep; these did not change from periods A to C. We conclude that children with OSAS have normal sleep stage distribution. OSAS is predominantly a REM phenomenon in children. Obstructive apnea worsens over the course of the night, independent of the changing amounts of REM sleep. We speculate that this increase in apnea severity may be secondary to upper airway muscle fatigue, changes in upper airway neuromotor control, or changes in REM density.

Apnea↗
PubMed1991

Snoring and sleep architecture.

The purpose of this study was to examine whether snoring adversely affects sleep architecture and sleep efficiency, and thus may account for the frequent complaints of daytime tiredness and fatigue expressed by heavy snorers. We recruited eight self-confessed heavy snorers and six self-confessed nonsnorers. All subjects had full nocturnal polysomnography, including continuous monitoring of snoring, which was quantified by counting the number of snores per hour of sleep (snoring index), the number of snores per minute of snoring time (snoring frequency), maximal and mean nocturnal sound intensity (dBmax and dBmean, respectively). We found that even the self-confessed nonsnorers snored lightly, with significantly smaller frequency and index than the heavy snorers. Sleep architecture was similar in both groups. Distribution of snoring among the sleep stages differed for light and heavy snorers: light snorers snored uniformly throughout all sleep stages, whereas heavy snorers tended to snore more during slow-wave and REM sleep. Snoring frequency and snoring index were similar during all sleep stages in light snorers, but they were higher during slow-wave sleep in heavy snorers. Wakefulness time after sleep onset and sleep efficiency correlated significantly with the snoring index. We conclude that although snoring does not affect sleep architecture in general, it influences sleep efficiency and wakefulness time after sleep onset; this may have an adverse effect on daytime function of heavy snorers.

Adult↗
PubMed1999

Effect of mouth leak on effectiveness of nasal bilevel ventilatory assistance and sleep architecture.

Mouth leak is common during nasal ventilatory assistance, but its effects on ventilatory support and on sleep architecture are unknown. The acute effect of sealing the mouth on sleep architecture and transcutaneous carbon dioxide tension (Ptc,CO2) was tested in 9 patients (7 hypercapnic) on longterm nasal bilevel ventilation with symptomatic mouth leak. Patients slept with nasal bilevel ventilation at their usual settings on two nights in random order. On one night, the mouth was taped closed. Leak was measured with a pneumotachograph. Median leak fell from 0.35+/-0.07 (mean +/- SEM) L x s(-1) untaped to 0.06+/-0.03 L x s(-1) taped. Ptc,CO2 fell in 8/9, including all hypercapnic patients. Across all patients, the mean Ptc,CO2 fell by 1.02+/-0.28 kPa (7.7+/-2.1 mm Hg) with taping (p = 0.007). Arousal index fell in every patient. Mean arousal index fell from 35.0+/-3.0 to 13.9+/-1.2 h(-1) (p<0.0001), and rapid eye movement (REM) sleep increased from 12.9+/-1.5% to 21.1+/-1.8% sleep time (p = 0.0016). Slow wave sleep changed inconsistently, from a mean of 13.1+/-1.6% to 19.5+/-2.2% of sleep (p = 0.09). Sleep latency and efficiency were unchanged. In four healthy volunteers ventilator-induced awake hypopharyngeal pressure swing during timed bilevel ventilation fell by 35+/-5% L(-1) x s(-1) of voluntary mouth leak (p<0.0001). Mouth leak reduces effective nasal bilevel ventilatory support, increases transcutaneous carbon dioxide tension, and disrupts sleep architecture.

Aged↗
PubMed2003

Optimization of neural network architecture using genetic programming improves detection and modeling of gene-gene interactions in studies of human diseases.

BACKGROUND: Appropriate definition of neural network architecture prior to data analysis is crucial for successful data mining. This can be challenging when the underlying model of the data is unknown. The goal of this study was to determine whether optimizing neural network architecture using genetic programming as a machine learning strategy would improve the ability of neural networks to model and detect nonlinear interactions among genes in studies of common human diseases. RESULTS: Using simulated data, we show that a genetic programming optimized neural network approach is able to model gene-gene interactions as well as a traditional back propagation neural network. Furthermore, the genetic programming optimized neural network is better than the traditional back propagation neural network approach in terms of predictive ability and power to detect gene-gene interactions when non-functional polymorphisms are present. CONCLUSION: This study suggests that a machine learning strategy for optimizing neural network architecture may be preferable to traditional trial-and-error approaches for the identification and characterization of gene-gene interactions in common, complex human diseases.

Algorithms↗