SEARCH · PubMed Health

Results for “Performance benchmarking”

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 1,027 records · Page 57Linked to original sources

PubMed1990

Computational neural networks: enhancing supervised learning algorithms via self-organization.

A neural network processing scheme is proposed which utilizes a self-organizing Kohonen feature map as the front end to a feedforward classifier network. The results of a series of benchmarking studies based upon artificial statistical pattern recognition tasks indicate that the proposed architecture performs significantly better than conventional feedforward classifier networks when the decision regions are disjoint. This is attributed to the fact that the self-organization process allows internal units in the succeeding classifier network to be sensitive to a specific set of features in the input space at the outset of training.

Algorithms↗

PubMed2005

A graphical model for estimating stimulus-evoked brain responses from magnetoencephalography data with large background brain activity.

This paper formulates a novel probabilistic graphical model for noisy stimulus-evoked MEG and EEG sensor data obtained in the presence of large background brain activity. The model describes the observed data in terms of unobserved evoked and background factors with additive sensor noise. We present an expectation maximization (EM) algorithm that estimates the model parameters from data. Using the model, the algorithm cleans the stimulus-evoked data by removing interference from background factors and noise artifacts and separates those data into contributions from independent factors. We demonstrate on real and simulated data that the algorithm outperforms benchmark methods for denoising and separation. We also show that the algorithm improves the performance of localization with beamforming algorithms.

Algorithms↗

PubMed2003

Thermostable continuous coupled assay for measuring glucose using glucokinase and glucose-6-phosphate dehydrogenase from the marine hyperthermophile Thermotoga maritima.

A novel, thermostable adaptation of the coupled-enzyme assay for monitoring glucose concentrations was developed for an optimal temperature of 85 degrees C. This is the first report of a thermostable glucostat from a marine hyperthermophile. The continuous assay, using glucokinase (Glk) and glucose-6-phosphate dehydrogenase (Gpd) from Thermotoga maritima, demonstrated robust activity over a range of temperatures (75-90 degrees C) and pH values (6.8- 8.5). Purified glucokinase had a monomeric molecular mass of 33.8kDa while that of glucose-6-phosphate dehydrogenase (D-glucose 6-phosphate:NADP oxidoreductase) was 57.5kDa. The high-temperature assay provided a method for directly assaying the activity of another hyperthermophilic enzyme, 1,4-beta-D-glucan glucohydrolase (GghA) from Thermotoga neapolitana. To provide a benchmark for protein-engineering experiments involving GghA, a three-enzyme continuous assay (performed at 85 degrees C), linking wild-type GghA, Glk, and Gpd, measured glucose produced from GghA's hydrolysis of cellobiose, one of GghA's secondary substrates. The assay established the kinetic behavior of wild-type GghA toward cellobiose and was used to screen for changes in the catalytic efficiency of variant GghA(s) induced by random mutagenesis. The assay's development will allow high-throughput screening of other thermostable glucose-producing enzymes, including those applicable to commercial biomass conversion.

Electrophoresis, Polyacrylamide Gel↗

PubMed2006

Limitations of a PI controller for a first-order nonlinear process with dead time.

A nonlinear process, the conical tank level process whose parameters vary with respect to the process variable, is considered. The time constant and gain of the chosen process vary as a function of level. A deliberate dead time is introduced in the system by delaying the valve opening. The limitations of the conventional PI controller tuned using Ziegler-Nichols settings for the chosen process are brought out. The servo and regulatory responses through simulation and experimentation for various magnitudes of set-point changes and load changes at various operating points with the controller tuned only at a chosen nominal operating point are obtained and analyzed. Regulatory responses for output load changes are studied. A benchmark system is presented to understand clearly how variations of system parameters affect the performance of the controller. A gain scheduled controller is proposed to handle the control problem. Present work illustrates the simplicity and effectiveness of the gain scheduling.

Journal Article↗

PubMed1998

Infection surveillance in home care: device-related incidence rates.

BACKGROUND: Four home health agencies of different sizes, including a rural agency with many branch offices, participated in a surveillance study. METHODS: These agencies used the same definitions for symptomatic urinary tract infections in patients with urinary catheters and for bloodstream infections in patients receiving intravenous therapy. Identical data and methods were used for calculating infection rates, with device-days consistently used as the denominator. Each agency's primary goal was to develop baseline information about its own infection rates and to use these statistics to detect problems and trends. Chart audits were performed if trends or problems were identified. A secondary goal was to compare or benchmark infection rates, because there is almost no published information in this area of home care. RESULTS: We found that our mean infection rates for symptomatic urinary tract infection among patients with urinary catheters and for bloodstream infection among patients receiving intravenous therapy were similar among the four agencies. The mean rates for all four agencies were 4.5 symptomatic urinary tract infections/1000 device-days and 1.1 bloodstream infections/1000 device-days. CONCLUSION: The methods used here can be implemented in other home care agencies. Using the same definitions, collecting data for the same type of infections, and using the same rate calculations make infection control benchmarking possible between home health agencies.

Home Care Services↗

PubMed2000

A connectionist model of complacency and adaptive recovery under automation.

Automation is intended to reduce the demands on operators in complex environments, thereby enhancing overall system performance. Although automation usually reduces workload, it is often accompanied by a decline in monitoring performance, an effect known as complacency. The circumstances under which complacency occurs and how it can be prevented, for example by intermittently returning control to the operator, are empirically well understood. To date, that empirical knowledge has not been accompanied by strong psychological theory. This article presents a computational model of human performance under automation based on connectionist principles. The model is shown to explain several benchmark findings, among them the basic complacency effect; the effect of the variability of automation reliability on complacency; the effect of task complexity; and the effect of intermittently returning control to the operator.

Attention↗

PubMed2006

The simulation of the three-dimensional lattice hydrophobic-polar protein folding.

One of the most prominent problems in computational biology is to predict the natural conformation of a protein from its amino acid sequence. This paper focuses on the three-dimensional hydrophobic-polar (HP) lattice model of this problem. The modified elastic net (EN) algorithm is applied to solve this nonlinear programming hard problem. The lattice partition strategy and two local search methods (LS(1) and LS(2)) are proposed to improve the performance of the modified EN algorithm. The computation and analysis of 12 HP standard benchmark instances are also involved in this paper. The results indicate that the hybrid of modified EN algorithm, lattice partition strategy, and local search methods has a greater tendency to form a globular state than genetic algorithm does. The results of noncompact model are more natural in comparison with that of compact model.

Algorithms↗

PubMed2006

Automated phylogenetic detection of recombination using a genetic algorithm.

The evolution of homologous sequences affected by recombination or gene conversion cannot be adequately explained by a single phylogenetic tree. Many tree-based methods for sequence analysis, for example, those used for detecting sites evolving nonneutrally, have been shown to fail if such phylogenetic incongruity is ignored. However, it may be possible to propose several phylogenies that can correctly model the evolution of nonrecombinant fragments. We propose a model-based framework that uses a genetic algorithm to search a multiple-sequence alignment for putative recombination break points, quantifies the level of support for their locations, and identifies sequences or clades involved in putative recombination events. The software implementation can be run quickly and efficiently in a distributed computing environment, and various components of the methods can be chosen for computational expediency or statistical rigor. We evaluate the performance of the new method on simulated alignments and on an array of published benchmark data sets. Finally, we demonstrate that prescreening alignments with our method allows one to analyze recombinant sequences for positive selection.

Algorithms↗

PubMed1992

Benchmark test of transport calculations of gold and nickel activation with implications for neutron kerma at Hiroshima.

A benchmark test of the Monte Carlo neutron and photon transport code system (MCNP) was performed using a 252Cf fission neutron source to validate the use of the code for the energy spectrum analyses of Hiroshima atomic bomb neutrons. Nuclear data libraries used in the Monte Carlo neutron and photon transport code calculation were ENDF/B-III, ENDF/B-IV, LASL-SUB, and ENDL-73. The neutron moderators used were granite (the main component of which is SiO2, with a small fraction of hydrogen), Newlight [polyethylene with 3.7% boron (natural)], ammonium chloride (NH4Cl), and water (H2O). Each moderator was 65 cm thick. The neutron detectors were gold and nickel foils, which were used to detect thermal and epithermal neutrons (4.9 eV) and fast neutrons (> 0.5 MeV), respectively. Measured activity data from neutron-irradiated gold and nickel foils in these moderators decreased to about 1/1,000th or 1/10,000th, which correspond to about 1,500 m ground distance from the hypocenter in Hiroshima. For both gold and nickel detectors, the measured activities and the calculated values agreed within 10%. The slopes of the depth-yield relations in each moderator, except granite, were similar for neutrons detected by the gold and nickel foils. From the results of these studies, the Monte Carlo neutron and photon transport code was verified to be accurate enough for use with the elements hydrogen, carbon, nitrogen, oxygen, silicon, chlorine, and cadmium, and for the incident 252Cf fission spectrum neutrons.

Californium↗

PubMed1994

Benchmark test of neutron transport calculations: indium, nickel, gold, europium, and cobalt activation with and without energy moderated fission neutrons by iron simulating the Hiroshima atomic bomb casing.

A benchmark test of the Monte Carlo neutron and photon transport code system (MCNP) was performed using a bare- and energy-moderated 252Cf fission neutron source which was obtained by transmission through 10-cm-thick iron. An iron plate was used to simulate the effect of the Hiroshima atomic bomb casing. This test includes the activation of indium and nickel for fast neutrons and gold, europium, and cobalt for thermal and epithermal neutrons, which were inserted in the moderators. The latter two activations are also to validate 152Eu and 60Co activity data obtained from the atomic bomb-exposed specimens collected at Hiroshima and Nagasaki, Japan. The neutron moderators used were Lucite and Nylon 6 and the total thickness of each moderator was 60 cm or 65 cm. Measured activity data (reaction yield) of the neutron-irradiated detectors in these moderators decreased to about 1/1,000th or 1/10,000th, which corresponds to about 1,500 m ground distance from the hypocenter in Hiroshima. For all of the indium, nickel, and gold activity data, the measured and calculated values agreed within 25%, and the corresponding values for europium and cobalt were within 40%. From this study, the MCNP code was found to be accurate enough for the bare- and energy-moderated 252Cf neutron activation calculations of these elements using moderators containing hydrogen, carbon, nitrogen, and oxygen.

Cobalt↗

PubMed2002

Developing high-performance leaders.

Although there is widespread recognition that strong leadership is key in these challenging times, many companies provide only the tip of the iceberg of leadership development support. This article is a resource for high-powered leadership development systems that will have an impact on performance. Four topics are discussed: (1) models, (2) investment and results, (3) critical success factors, and (4) case studies of how the 3M Company and HealthPartners develop high-performance leaders. Studies that quantity the effect of leadership development on performance are noted. Five critical success factors are described, and examples from leadership development benchmark organizations such as General Electric and Reell Precision Manufacturing are discussed.

Benchmarking↗

PubMed2023

Unraveling Neuronal Identities Using SIMS: A Deep Learning Label Transfer Tool for Single-Cell RNA Sequencing Analysis.

Large single-cell RNA datasets have contributed to unprecedented biological insight. Often, these take the form of cell atlases and serve as a reference for automating cell labeling of newly sequenced samples. Yet, classification algorithms have lacked the capacity to accurately annotate cells, particularly in complex datasets. Here we present SIMS (Scalable, Interpretable Machine Learning for Single-Cell), an end-to-end data-efficient machine learning pipeline for discrete classification of single-cell data that can be applied to new datasets with minimal coding. We benchmarked SIMS against common single-cell label transfer tools and demonstrated that it performs as well or better than state of the art algorithms. We then use SIMS to classify cells in one of the most complex tissues: the brain. We show that SIMS classifies cells of the adult cerebral cortex and hippocampus at a remarkably high accuracy. This accuracy is maintained in trans-sample label transfers of the adult human cerebral cortex. We then apply SIMS to classify cells in the developing brain and demonstrate a high level of accuracy at predicting neuronal subtypes, even in periods of fate refinement, shedding light on genetic changes affecting specific cell types across development. Finally, we apply SIMS to single cell datasets of cortical organoids to predict cell identities and unveil genetic variations between cell lines. SIMS identifies cell-line differences and misannotated cell lineages in human cortical organoids derived from different pluripotent stem cell lines. When cell types are obscured by stress signals, label transfer from primary tissue improves the accuracy of cortical organoid annotations, serving as a reliable ground truth. Altogether, we show that SIMS is a versatile and robust tool for cell-type classification from single-cell datasets.

Brain organoids↗

PubMed2004

Improving generalization performance of natural gradient learning using optimized regularization by NIC.

Natural gradient learning is known to be efficient in escaping plateau, which is a main cause of the slow learning speed of neural networks. The adaptive natural gradient learning method for practical implementation also has been developed, and its advantage in real-world problems has been confirmed. In this letter, we deal with the generalization performances of the natural gradient method. Since natural gradient learning makes parameters fit to training data quickly,the overfitting phenomenon may easily occur, which results in poor generalization performance. To solve the problem, we introduce the regularization term in natural gradient learning and propose an efficient optimizing method for the scale of regularization by using a generalized Akaike information criterion (network information criterion). We discuss the properties of the optimized regularization strength by NIC through theoretical analysis as well as computer simulations. We confirm the computational efficiency and generalization performance of the proposed method in real-world applications through computational experiments on benchmark problems.

Algorithms↗

PubMed2005

Evolving evolutionary algorithms using linear genetic programming.

A new model for evolving Evolutionary Algorithms is proposed in this paper. The model is based on the Linear Genetic Programming (LGP) technique. Every LGP chromosome encodes an EA which is used for solving a particular problem. Several Evolutionary Algorithms for function optimization, the Traveling Salesman Problem and the Quadratic Assignment Problem are evolved by using the considered model. Numerical experiments show that the evolved Evolutionary Algorithms perform similarly and sometimes even better than standard approaches for several well-known benchmarking problems.

Algorithms↗

PubMed2006

Fine-needle aspiration biopsy of lymphoproliferative disorders--interpretations based on morphologic criteria alone: results from the College of American Pathologists Interlaboratory Comparison Program in Nongynecologic Cytopathology.

CONTEXT: Diagnosis of lymphoproliferative disorders is one of the most challenging tasks faced by the cytologist. The initial cytomorphologic evaluation of lymphoproliferative lesions directs the choice of ancillary studies that ultimately lead to a diagnosis based on the World Health Organization classification system using a composite of clinical, morphologic, immunophenotypic, and molecular features. OBJECTIVE: To evaluate the ability of participating laboratories in the College of American Pathologists Interlaboratory Comparison Program in Non-Gynecologic Cytopathology to appropriately categorize lymphoproliferative lesions based solely on cytomorphologic criteria. DESIGN: Laboratory responses for lymph node aspirates were examined. All responses were based on review of glass slides without ancillary immunologic or molecular data available. The benchmarking data provided for each specific diagnosis were analyzed, with a focus on the performance for evaluation of lymphoproliferative lesions. RESULTS: Based on morphology alone, responses for lymph node aspirates in the Non-Gynecologic Cytopathology program were correct to the exact reference diagnosis for 87.1% of Hodgkin lymphoma. Non-Hodgkin lymphoma was identified in 69.5% of the large cell non-Hodgkin lymphoma cases, of which 66.8% were correctly classified as large cell type. Non-Hodgkin lymphoma was identified in 68.1% of non-Hodgkin lymphoma, other than large cell cases, and of these, 94.7% were identified as other than large cell type. CONCLUSIONS: The spectrum of specific responses was consistent for lymphoproliferative lesions, with a reasonable differential diagnosis based on cytomorphology alone, which, in practice, facilitates the appropriate choice of immunophenotypic markers and other ancillary studies.

Biopsy, Fine-Needle↗

PubMed1997

Clinical performance measures focus on process steps.

Performance measures go beyond mortality and cost: Using process steps as performance measures allows providers to easily determine where action needs to be taken to improve quality. One consultant has developed a list of 37 performance measures in nine disease categories and tracks performance using a software program. Here are the details, plus some disease-specific benchmarks.

Benchmarking↗

PubMed1997

Spinal shrinkage during work in a sitting posture compared to work in a standing posture.

OBJECTIVE: The objective was to measure the possible differences in shrinkage of the thoracolumbar spine in subjects working in a sitting and a standing posture for 6.5 h at work, in a realistic work environment. The isolated shrinkage of the thoracic and the lumbar spine was also examined. STUDY DESIGN: This study presents a new protocol to measure shrinkage of the thoracic and lumbar spine separately. BACKGROUND: Controversies still exist with regard to the load on the spine in a sitting compared to a standing position. Some report that shrinkage is greatest in the sitting position while others report the reverse. However, nothing is known about the height reduction of the thoracic and the lumbar spine during loading for 6.5 h in a real work environment. Therefore, the behaviour of the thoracic and the lumbar spine under practical condition has to be investigated. METHOD: A stadiometer with a measurement error of 0.51 mm was used to measure changes in spinal height during work. To exclude first-time behaviour of the spine, a pre-test lasting 50 min was undertaken. The mean of the last three measurements was used as the reference height. During work, height measurements of the spine were performed every 20 min. To separate the behaviour of the thoracic and the lumbar spine, two benchmarks were placed at the vertebrae prominens and at the thoracic-lumbar junction. Shrinkage of the spine was investigated within three different cohorts: (I) work in a sitting posture for 6.5 h; (II) relaxed sitting for 2 h vs work for 2 h in a sitting position and (III) work in a standing position for 6.5 h. RESULTS: Relaxed sitting leads to a gain in stature compared to work in a sitting position for 2 h. The major gain in stature occurred in the lumbar spine. Comparison of cohort (III) working in a standing position with cohort (I) working in a sitting position shows that the shrinkage of the spine is greatest when work is performed in a standing posture. The major differences were found in the shrinkage of the lumbar spine, e.g. shrinkage of the lumbar spine in the standing cohort (III) was 4.16 mm compared to 1.73 mm in the sitting cohort (I). CONCLUSIONS: There is a gain in stature during relaxed sitting compared to work in a sitting posture. The load on the spine is greatest when work in a standing position is performed. The greater shrinkage of the lumbar spine during work in a standing position compared to a sitting posture is probably due to: (i) differences in lumbar lordosis and (ii) the effect of bending and torsion while handling the work materials.

Journal Article↗

PubMed1995

Adrenalectomy via the dorsal approach: a benchmark for laparoscopic adrenalectomy.

PURPOSE: We reviewed our experience with posterior surgical excision of aldosteronomas. MATERIALS AND METHODS: A retrospective review identified 40 consecutive cases performed by 1 surgeon. RESULTS: Mean patient age was 46 years and mean blood loss was 237 cc. There was no perioperative mortality or intraoperative complications. Times to unassisted ambulation and return to a normal diet were 2.2 and 2.4 days, respectively, with a mean postoperative hospital stay of 4.4 days. CONCLUSIONS: The removal of aldosteronomas via the dorsal approach is a safe reliable technique with acceptable morbidity. These results provide a benchmark by which the results of laparoscopic adrenalectomy may be judged.

Adrenal Cortex Neoplasms↗