PubMed Health⌕ Search

SEARCH · PubMed Health

Results for “Error minimization”

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 505 records · Page 28Linked to original sources
PubMed2006

Tradeoffs of different types of species occurrence data for use in systematic conservation planning.

Data on the occurrence of species are widely used to inform the design of reserve networks. These data contain commission errors (when a species is mistakenly thought to be present) and omission errors (when a species is mistakenly thought to be absent), and the rates of the two types of error are inversely related. Point locality data can minimize commission errors, but those obtained from museum collections are generally sparse, suffer from substantial spatial bias and contain large omission errors. Geographic ranges generate large commission errors because they assume homogenous species distributions. Predicted distribution data make explicit inferences on species occurrence and their commission and omission errors depend on model structure, on the omission of variables that determine species distribution and on data resolution. Omission errors lead to identifying networks of areas for conservation action that are smaller than required and centred on known species occurrences, thus affecting the comprehensiveness, representativeness and efficiency of selected areas. Commission errors lead to selecting areas not relevant to conservation, thus affecting the representativeness and adequacy of reserve networks. Conservation plans should include an estimation of commission and omission errors in underlying species data and explicitly use this information to influence conservation planning outcomes.

Animals↗
PubMed1998

Verification of set-up deviations in patients with breast cancer using portal imaging in clinical practice.

AIM: The aim of this study was to determine set-up deviations during irradiation of patients with breast cancer using Electronic Portal Imaging (EPI). PATIENTS AND METHOD: In order to monitor the intrafractional set-up deviations multiple portal image readings were carried out on 5 patients with breast cancer. Moreover interfractional set-up deviations were determined by the acquisition of daily EPIs in a total of 20 patients. RESULTS: The results of intrafractional set-up deviation based on 130 EPIs revealed a maximum lateral and longitudinal (cranio-caudal) deviation of 2 mm (range:-10 mm to 8 mm) as well as 1 degree in rotation (range: -2 degree to 2 degrees). The interfractional set-up deviations in 20 breast cancer patients during a treatment series of 25 fractions showed mean standard deviations of 5 mm in lateral and longitudinal direction, respectively. Only in few cases deviations up to maximally 24 mm were observed. The mean standard deviation of the rotational error was 2 degrees and reached a maximum of 6.5 degrees. CONCLUSION: These results show that intrafractional set-up deviations in breast cancer patients are negligible in clinical practice. They can be attributed to random errors due to patient movement and breathing. The set-up deviations during a treatment series can be differentiated in systematic and random errors. Patient fixation and immobilization is crucial in minimizing random errors. Taking into account 10 mm safety margins as used in our department around the "clinical target volume" (CTV), set-up errors outside of this volume (PTV) were exceptional. The enlarged PTV definition of 15 mm for lateral and cranio-caudal field margins, respectively as used in this study covers 99% of the CTV in all patients.

Breast Neoplasms↗
PubMed2004

FloatBoost learning and statistical face detection.

A novel learning procedure, called FloatBoost, is proposed for learning a boosted classifier for achieving the minimum error rate. FloatBoost learning uses a backtrack mechanism after each iteration of AdaBoost learning to minimize the error rate directly, rather than minimizing an exponential function of the margin as in the traditional AdaBoost algorithms. A second contribution of the paper is a novel statistical model for learning best weak classifiers using a stagewise approximation of the posterior probability. These novel techniques lead to a classifier which requires fewer weak classifiers than AdaBoost yet achieves lower error rates in both training and testing, as demonstrated by extensive experiments. Applied to face detection, the FloatBoost learning method, together with a proposed detector pyramid architecture, leads to the first real-time multiview face detection system reported.

Algorithms↗
PubMed2005

Error optimization of a three-dimensional magnetic resonance imaging tagging-based cartilage deformation technique.

Three-dimensional strain fields in articular cartilage subjected to compressive loading can be determined using a recently developed MRI-based cartilage deformation by tag registration technique. The objective of this study was to determine the experimental variables that minimize the technique error, which has not been previously reported. Error (strain bias and precision) was determined using direct experiments and Monte Carlo simulations for four variables: spatial resolution, tag line spacing, applied nominal strain, and number of control points used to describe tag lines in a B-spline model. The important results include the following: (1) bias was not significantly different from zero, (2) precision increased with image resolution and with tag line spacing, (3) precision was independent of applied nominal strain, and (4) error was a minimum (absolute precision = 0.41% strain) for the following values: spatial resolution = 0.05 x 0.05 mm2; tag line spacing = 2.0 mm; control points = 6. With these results the technique can now be used in various applications while minimizing error.

Algorithms↗
PubMed2001

Incorporation of loudness measures in active noise control.

An attempt has been made to use a modified version of a standard active noise control algorithm in order to take into account the unique response of the human auditory system. It has been shown in the past that decreasing the sound pressure level at a location does not guarantee a similar decrease in the perceived loudness at that location. Typically, active noise control is based on minimizing the "error signal" from a mechanical device such as a microphone, whose response is nominally flat across the frequency response range of the human ear. However, if the response of the ear can be approximated by digitally filtering the error signal before it reaches the adaptive controller, one can, in effect, minimize the more subjective loudness level, as opposed to the sound pressure level. The work reported here entails simulating active noise control based upon minimizing perceived loudness for a collection of input noise signals. A comparison of the loudness of the resulting error signal is made to the loudness of that resulting from standard sound pressure level minimization. It has been found that the effectiveness of this technique is largely dependent upon the nature of the input noise signal. Furthermore, this technique is judged to be worth considering for use with applications of active noise control where the uncontrolled noise more prominently constitutes low range audio frequencies (approximately 30 Hz-100 Hz) than medium range audio frequencies (approximately 300 Hz-600 Hz).

Algorithms↗
PubMed1997

Comparison of anti-factor Xa heparin activity and activated partial thromboplastin time in 2,773 plasma samples from unfractionated heparin-treated patients.

This study was conducted to evaluate two methods of determining the "therapeutic range" of the activated partial thromboplastin time (APTT) during unfractionated heparin treatment. Multiple fresh plasma samples from 694 patients treated with unfractionated heparin were tested simultaneously for APTT and anti-factor Xa heparin activity. The data were analyzed by linear regression and by a minimization-of-error technique to determine the more accurate APTT therapeutic range. The best-fit linear regression equation was obtained using logarithmically transformed APTT values. Using this equation, the heparin activity predicted by an APTT value had a 95% limit of agreement of +/- 0.39 U/mL compared with the actual heparin activity. The total clinical error rate of the APTT therapeutic range selected using the best-fit regression equation was 10%. An improvement to 8% was achieved using a method that chose the therapeutic range based on minimizing the APTT errors. Even 8% may be too high an error rate for best clinical results.

Evaluation Studies as Topic↗
PubMed1998

Integrated software for a macromolecular crystallography synchrotron beamline.

A package of software has been produced for the operation of the synchrotron beamline X12-C at the National Synchrotron Light Source at Brookhaven National Laboratory. Years of observing common user mistakes has enabled the production of software that reduces user errors significantly. Users of the beamline communicate with all the experimental apparatus, including both the data-collection equipment and the beamline components, including the monochromator, through an easy-to-use graphical user interface (GUI). Important features of the system are (1) its modularity, so that different underlying programs or different apparatus can be incorporated easily; (2) its ease of use, minimizing both user errors and training effort; and (3) that most of the experimental operations and parameters are logged automatically, again minimizing errors and facilitating more-or-less automatic reduction of the data. Features of the software are useful enough to have been incorporated into operations at other synchrotron beamlines.

Crystallography, X-Ray↗
PubMed1992

A microcomputer program for evaluating sampling error: an application to stereological methods for electron microscopy.

In situations where there is a need to minimize sampling error or sample size, the coefficient of variation (CV) may be used to evaluate sampling error as a function of the number of observations or subjects in a sample. For example, CV is useful for estimating the minimum number of electron micrographs (Nmin) required to obtain a representative field sample for stereological analysis. To facilitate the determination of Nmin, we have written a program (COEFficient) for DOS microcomputers which calculates CVs. COEF assists the user in reducing error to that which solely reflects biological variability, thereby minimizing the time and cost of subsequent analyses.

Microcomputers↗
PubMed2002

Error correction strategies for motor behavior after unilateral brain damage: short-term motor learning processes.

In order to identify the mechanisms underlying motor impairments and motor learning following stroke-related brain damage, we analyzed correction strategies used by hemiparetic individuals to produce precise elbow flexion movements of the paretic arm and compared them to those of healthy individuals. Participants made rapid elbow flexion movements to a 6 degrees wide target and were instructed to correct movement errors as quickly as possible when a spring-like load was unexpectedly introduced. Angular positions and torques before correction were used to identify error patterns. Results showed that participants with mild hemiparesis minimized movement errors within three trials, as did healthy participants. In contrast, severely affected individuals needed more trials to diminish errors and their movements were inconsistent. Participants with a moderate motor disability used both typical and atypical correction strategies. The differences in correction behaviors likely reflect deficits in arm motor function (r=0.79) and executive function (r=0.58) rather than levels of intellectual function (IQ ratings). Results indicate that the deficits that individuals with stroke experience when adapting their movements to changed load conditions may be due to difficulty in rapidly integrating visual and proprioceptive information. Deficits in executive function could also contribute to problems in producing accurate and consistent movements from trial to trial. Taken together, these results imply that all hemiparetic individuals would not benefit equally from the same motor re-training approaches.

Adolescent↗
PubMed2001

Error correcting memorization learning for noisy training examples.

In order to avoid overfitting, we propose error correcting memorization learning. This method is derived from minimization of error between outputs of a trained neural network and correct values for noisy training examples, although the correct values are unknown. We show that noise is adequately suppressed by error correcting memorization learning. The noise suppression mechanism is theoretically clarified. It is found that redundancy plays an essential role for noise suppression and depends on a set of training inputs. We give the condition for the training inputs to provide the redundancy. Moreover, by clarifying the relationships between the proposed method and the weighted least squares estimation with the Mahalanobis norm, we reveal effectiveness of the weighted least squares estimation on noise suppression.

Models, Theoretical↗
PubMed2003

The neuropsychological effects of obstructive sleep apnea: a meta-analysis of norm-referenced and case-controlled data.

STUDY OBJECTIVES: The research literature on the neuropsychological effects of obstructive sleep apnea (OSA) has yielded seemingly contradictory findings, and narrative reviews of this literature are prone to interpretive errors. We used sophisticated meta-analytic models to minimize such errors, with the goal of clarifying the effect of OSA on neuropsychological functioning. DESIGN: Meta-analytic review of research through 2001. PARTICIPANTS: We reviewed studies of neuropsychological functioning among adults with untreated OSA. Twenty-five studies met review criteria, representing 1092 patients with OSA and 899 healthy controls. MEASUREMENTS AND RESULTS: Two sets of effect sizes were generated. One compared OSA group means against those of healthy controls in case-controlled studies. The other compared all OSA group means against published normative data. Within each data set, 10 neuropsychological outcome domains were coded. In both data sets, untreated OSA was found to have a negligible impact on intellectual and verbal functioning but a substantial impact upon vigilance and executive functioning. Data were mixed with regard to visual and motor functioning; posthoc inspection of the data suggested that tests of fine-motor coordination or drawing were more sensitive to OSA than were tests of fine-motor speed or visual perception. Data were also mixed with regard to memory functioning, probably related to methodologic differences across studies. CONCLUSIONS: Etiologic models should emphasize mechanisms known to affect vigilance, executive functioning, and motor coordination but not intelligence, verbal functioning, or visual perception. Clinicians should be alert to OSA symptoms in patients with declines in vigilance, executive functioning, or coordination.

Arousal↗
PubMed2001

Force transducer design: a new approach combining nonlinear finite element analysis and robust design.

Designing a strain gage based force transducer is considered more of an art than an engineering process. Only general guidelines are available, and many trial and error iterations are needed to optimize the geometry and minimize the errors caused by nonlinear behavior of the structure. A new method, based on nonlinear finite element analysis and robust design principles, is proposed. A matrix of experiments considers relevant geometric and loading parameters. The behavior of the structure under different combinations of these parameters is determined by calculating the strain at different locations (suitable for strain gage installation) using finite element models. The nonlinear behavior of the transducer is identified by comparing the results of the nonlinear finite element analysis with those obtained using a linear finite element analysis. A signal-to-noise ratio is defined to quantify the nonlinearities and how the considered parameters affect them. An analysis of variance is employed to determine their relative influence. Based on the results of the statistical analysis, it is possible to identify the best value for each geometric parameter that would reduce, if not eliminate the nonlinearities. Once these optimal geometric parameters are chosen, a prototype can be built, instrumented with strain gages, and tested, to validate the obtained design. To illustrate this new proposed methodology, and appreciate its advantages over current practice of designing a force transducer, an example of the step-by-step procedure is illustrated considering a thin-wall cylindrical transducer.

Equipment Design↗
PubMed1983

Factors affecting the error in the Loo-Riegelman method for estimating the rate of drug absorption. Some suggestions for a practical sampling schedule.

During the development of a new computer program for the Loo-Riegelman method for the estimation of the rate drug absorption, it was found that many factors can affect the accuracy of the estimated absorption rate constant (ka). In this paper a survey of the factors affecting the errors in the estimation of ka by the Loo-Riegelman method is made. By simulating data and computer techniques, the effects of factors related to each step in the calculation procedure as well as the influence of certain pharmacokinetic parameters were studied in detail. If these factors are not well controlled, the estimated error in ka will become very large, in some cases even surpassing 30%. In addition, it was derived that careful attention has to be paid to both sampling time and the number of plasma samples. A "trial and error" method for obtaining an optimal sampling schedule that minimizes the error of ka is presented.

Intestinal Absorption↗
PubMed2000

Minimal-model estimates of insulin sensitivity are insensitive to errors in glucose effectiveness.

The minimal-model method allows for estimation of insulin sensitivity (S(I) = P(3)/P(2)) and glucose effectiveness (S(G) = P(1)) from the time course of glucose and insulin after a glucose bolus. We previously demonstrated that the minimal-model results in overestimates of S(G) in subjects with normal insulin secretory function. To determine whether overestimation of S(G) has an impact on estimation of S(I), we examined model estimation of S(I) when S(G) was constrained to levels below that found by the regular minimal-model fit. Fifty-six glucose tolerance tests from lean and obese women, with and without polycystic ovary syndrome, were used. S(I) ranged from 0.2-22.6 x 10(-4) min(-1)/(microU/mL), and S(G) ranged from 0.8-3.8 x 10(-2) min(-1) for the standard minimal-model fits. Constraining S(G) to as low as 40% of the unconstrained value resulted in a 4-fold increase in P2 and P3, but only a 3% reduction in S(I). We conclude that estimation of the insulin sensitivity index is independent of errors in minimal-model-derived estimates of glucose effectiveness.

Adult↗
PubMed2005

Improving the accuracy and reliability of motion tracking methods used for extracting temporal motor activity signals from video recordings of neonatal seizures.

This paper presents an approach for improving the accuracy and reliability of motion tracking methods developed for video based on block motion models. This approach estimates the displacement of a block of pixels between two successive frames by minimizing an error function defined in terms of the pixel intensities at these frames. The minimization problem is made analytically tractable by approximating the error function using a second-order Taylor expansion. The improved reliability of the proposed method is illustrated by its application in the extraction of temporal motor activity signals from video recordings of neonatal seizures.

Algorithms↗
PubMed1990

Accuracy of four indirect methods of blood pressure measurement, with hemodynamic correlations.

In 38 adults undergoing cardiac surgery, 4 indirect blood pressure techniques were compared with brachial arterial blood pressure at predetermined intervals before and after cardiopulmonary bypass. Indirect blood pressure measurement techniques included automated oscillometry, manual auscultation, visual onset of oscillation (flicker) and return-to-flow methods. Hemodynamic measurements or calculations included heart rate, cardiac index, stroke volume index, and systemic vascular resistance index. Indirect and intraarterial blood pressure values were compared by simple linear regression by patient and measurement period. Measurement errors (arterial minus indirect blood pressure) were calculated, and stepwise regression assessed the relationship between measurement error and heart rate, cardiac index, stroke volume index, and systemic vascular resistance index. Indirect to intraarterial blood pressure correlation coefficients varied over time, with the strongest correlations often occurring at the first and last measurement periods (preinduction and 60 minutes after cardiopulmonary bypass), particularly for systolic blood pressure. Within-patient correlations between indirect and arterial blood pressure varied widely--they were consistently high or low in some patients. In other patients, correlations were especially weak with a particular indirect blood pressure method for systolic, mean, or diastolic blood pressure; in some cases indirect blood pressure was inadequate for clinical diagnosis of acute blood pressure changes or trends. The mean correlations between indirect and direct blood pressure values were, for systolic blood pressure: 0.69 for oscillometry, 0.77 for auscultation, 0.73 for flicker, and 0.74 for return-to-flow; for mean blood pressure: 0.70 for oscillometry and 0.73 for auscultation; and for diastolic blood pressure: 0.73 for oscillometry and 0.69 for auscultation. The mean measurement errors (arterial minus indirect values) for the individual indirect blood pressure methods were, for systolic: 0 mm Hg for oscillometry, 9 mm Hg for auscultation, -5 mm Hg for flicker, 7 mm Hg for return-to-flow; for mean: -6 mm Hg for oscillometry, and -3 mm Hg for auscultation; and for diastolic: -9 mm Hg for oscillometry and -8 mm Hg for auscultation. Mean measurement error for systolic blood pressure was thus least with automated oscillometry and greatest with manual auscultation, while standard deviations ranging from 9 to 15 mm Hg confirmed the highly variable nature of single indirect blood pressure measurements. Except for oscillometric diastolic blood pressure, a combination of systemic hemodynamics (heart rate, stroke volume index, systemic vascular resistance index, and cardiac index) correlated with each indirect blood pressure measurement error, which suggests that particular numeric ranges of these variables minimize measurement error.(ABSTRACT TRUNCATED AT 400 WORDS)

Adult↗
PubMed1998

Computational blood flow modelling: errors associated with reconstructing finite element models from magnetic resonance images.

Construction of computational blood flow models from magnetic resonance (MR) scans of real arteries is a powerful tool for studying arterial hemodynamics. In this report we experimentally determine a lower bound for errors associated with such an approach, and present techniques for minimizing such errors. A known, simple three-dimensional geometry (cylindrical tube) was imaged using a commercial MR scanner, and the resulting images were used to construct finite element flow models. Computed wall-shear stresses were compared to known values and peak errors of 40-60% were found. These errors can be attributed to limited spatial resolution, image segmentation and model construction. A simple smoothing technique markedly reduced these peak errors. We conclude that smoothing is required in the construction of arterial models from in vivo MR images. If used appropriately, such images can be used to construct acceptably accurate computational models of realistic arterial geometries.

Blood Circulation↗