SEARCH · PubMed Health

Results for “Statistics, Nonparametric”

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 199 records · Page 11Linked to original sources

PubMedDate not supplied

Localized spatial discrimination of epicardial conduction paths after linear transformation of variant information.

We present a method for the localized statistical discrimination of class populations based on the Karhunen-Loève and Fukunaga-Koontz transforms. These transforms provide features that model the variance of a sample distribution. The spatial series of a 196 channel epicardial electrogram recording from an arrythmogenic postinfarct canine were analyzed. For each type of rhythm studied, Karhunen-Loève and Fukunaga-Koontz expansions were computed from five training sets of spatial data, corresponding to five locations across the surface of the heart. Nonparametric statistical tests were then used for discriminant analysis to compare properties representative of the distribution from each proposed class. In a comparison of properties from sinus rhythm to those of two ventricular tachycardias, several spatial regions exhibited statistically significantly different propagation characteristics. These areas were observed by visual inspection of electrogram activation maps to be characterized by conductive gradients, which differed in magnitude and direction from one rhythm to another. The regions in which the propagation characteristics are of greatest difference in each tachycardia were centered upon sites of conduction block, manifested by reentrant circuit rhythms. Therefore, the importance of the technique for the localization of specific electrophysiologic events is demonstrated. This study extends previous work of our group on biosignal pattern recognition to encompass localized spatial data.

Analysis of Variance↗

PubMed1994

Parametric survival analysis of adjuvant therapy for stage II breast cancer.

BACKGROUND: Standard, nonparametric statistical methods estimate only the impact of therapy on survival rate up to a selected follow-up interval. In contrast, parametric methods can estimate the impact of treatment on the two cardinal parameters of malignancy: likelihood of cure and recurrence free survival time among uncured patients. METHODS: The authors screened a total of six parametric survival models. Three of these, including the log normal model, were applied to survival data from five clinical trials of adjuvant therapy for Stage II breast cancer. For comparison, the log rank test, a standard nonparametric method, was also applied to the same data. RESULTS: Both parametric and nonparametric methods identified a significant therapeutic in three of the five trials. In only one of these three trials, however, did parametric analysis identify a significant difference in the likelihood of cure between treatment groups. In the remaining two trials, a significant difference was found in recurrence free survival time among uncured patients. The three parametric survival models gave similar results. CONCLUSION: These findings suggest that parametric analysis may warrant further study as a method for measuring the long term clinical impact of adjuvant therapy on Stage II breast cancer.

Breast Neoplasms↗

PubMed2006

Propagation of interictal discharges in temporal lobe epilepsy: correlation of spatiotemporal mapping with intracranial foramen ovale electrode recordings.

OBJECTIVE: We have investigated intracerebral propagation of interictal epileptiform discharges (IED) in patients with mesial temporal lobe epilepsy (MTLE) by using spatiotemporal source maps based on statistical nonparametric mapping (SNPM) of low resolution electromagnetic tomography (LORETA) values. METHODS: We analyzed 30 patterns of IED recorded simultaneously with scalp and intracranial foramen ovale (FO) electrodes in 15 consecutive patients with intractable MTLE. The scalp EEG signals were averaged time-locked to the peak activity in bilateral 10-contact FO electrode recordings. SNPM was applied to LORETA values and spatiotemporal source maps were created by allocating the t-values over time to their corresponding Brodmann areas. Propagation was defined as secondary statistically significant involvement of distinct cortical areas separated by >15 ms. The results were correlated with intracranial data obtained from FO electrode recordings and with scalp EEG recordings. All patients underwent subsequent amygdalo-hippocampectomy and outcome was assessed one year after surgery. RESULTS: We found mesial to lateral propagation in 6/30 IED patterns (20%, four patients), lateral to mesial propagation in 4/30 IED patterns (13.3%, four patients) and simultaneous (within 15 ms) activation of mesial and lateral temporal areas in 6/30 IED patterns (20%, five patients). Propagation generally occurred within 30 ms and was always limited to ipsilateral cortical regions. Nine/30 IED patterns (30%) showed restricted activation of mesial temporal structures and no significant solutions were found in 5/30 IED patterns (16.7%). There was no clear association between the number or characteristics of IED patterns and the postsurgical outcome. CONCLUSIONS: Spatiotemporal mapping of SNPM LORETA accurately describes mesial to lateral temporal propagation of IED, and vice versa, which commonly occur in patients with MTLE. SIGNIFICANCE: Intracerebral propagation must be considered when using non-invasive source algorithms in patients with MTLE. Spatiotemporal mapping might be useful for visualizing this propagation.

Adult↗

PubMedDate not supplied

BASIC computer program to summarize data using nonparametric and parametric statistics including Anderson-Darling test for normality.

Tools for calculating Gaussian parametric statistics are widely available in hand-held calculators and program packages for desktop as well as mainframe computers. The standard deviation calculated from non-Gaussian data may, however, be meaningless or even absurd. Furthermore, the standard error of the mean is frequently used as a descriptive statistic even if it really does not describe the variation in the observations themselves. However, comprehensive tools for calculating nonparametric descriptive statistics and judging whether observations have a Gaussian distribution are hard to come by. The present paper attempts to provide the essential statistical reasoning behind the calculation of descriptive statistics and introduces a computer program written in Microsoft BASIC that applies the methods described in the text.

Mathematical Computing↗

PubMed2003

Ab-initio prediction and reliability of protein structural genomics by PROPAINOR algorithm.

We have formulated the ab-initio prediction of the 3D-structure of proteins as a probabilistic programming problem where the inter-residue 3D-distances are treated as random variables. Lower and upper bounds for these random variables and the corresponding probabilities are estimated by nonparametric statistical methods and knowledge-based heuristics. In this paper we focus on the probabilistic computation of the 3D-structure using these distance interval estimates. Validation of the predicted structures shows our method to be more accurate than other computational methods reported so far. Our method is also found to be computationally more efficient than other existing ab-initio structure prediction methods. Moreover, we provide a reliability index for the predicted structures too. Because of its computational simplicity and its applicability to any random sequence, our algorithm called PROPAINOR (PROtein structure Prediction by AI an Nonparametric Regression) has significant scope in computational protein structural genomics.

Algorithms↗

PubMed2002

Statistics review 6: Nonparametric methods.

The present review introduces nonparametric methods. Three of the more common nonparametric methods are described in detail, and the advantages and disadvantages of nonparametric versus parametric methods in general are discussed.

Humans↗

PubMedDate not supplied

Increasing efficiency and precision of data analysis: multivariate vs. univariate statistical techniques.

In deciding upon an appropriate analysis strategy during the planning phase of a research study, it is important to specify all of the independent and dependent variables to be included. From there, a technique should be chosen that will yield all of the information desired in the most efficient, precise, and powerful way. Frequently, the method of choice in nursing research will be a multivariate technique, since so many studies involve numerous variables whose effects on, or relationships with, other variables are of interest, as well as involving additional variables that should be taken into account for control or generalizability purposes. Not discussed in this paper, but worth mention, is the fact that all of these techniques involve various underlying assumptions (e.g., normality, homogeneity of variance, and independence in ANOVA) that must be met in order to use the techniques appropriately. If the assumptions are not met, the researcher might want to consider the use of nonparametric techniques--and there are both univariate and multivariate techniques available to choose from (Hollander & Wolfe, 1973; Siegel, 1956). The general advantages of a multivariate technique rather than separate univariate techniques would apply in terms of nonparametric statistics as well as parametric statistics. Increasing the accuracy, power, and efficiency of data analysis strategies should be a major concern to researchers.(ABSTRACT TRUNCATED AT 250 WORDS)

Analysis of Variance↗

PubMed1982

Statistical analysis and normality of selected hematologic and clinical chemistry measurements used in toxicologic studies.

The statistical comparison of hematological and clinical chemistry measurements of treated and control groups in toxicological studies has frequently, in the past, been done by methods that assumed, in the population of such values, normality of the data. This assumption was tested on pairs of concurrent control groups of rats from three long-term toxicological studies; two feeding and one inhalation routes of exposure were used. The normality of the data from three treated groups in one of these studies was also tested. The measurements examined were erythrocyte count, hemoglobin, white blood cell count, neutrophils, monocytes, urea nitrogen, and creatinine. In many cases significant skewness, kurtosis and deviation from normality as measured by the Kolmogorov-Smirnov test were present. The lack of consistency of normality versus nonnormality or, in some cases, direction of skewness, suggests that nonparametric statistical procedures be considered for use in the future for comparison of these types of data. Furthermore, the agreement of the data of the concurrent control groups as compared to the divergence of the values among the three studies affirms the importance of concurrent, in preference to historical, control data.

Animals↗

PubMed1997

CVLT recognition-recall discrepancies and retrieval deficits. A comment on Wilde et al (1995)

The methodological procedures employed by the authors are inappropriate for their data. Even when the results of the statistical analyses are taken at face value, the interpretation that there is no difference between the Discrepancy and Nondiscrepancy group is unsustainable, and the authors' conclusions that retrieval deficits do not underlie recognition-recall discrepancies is unwarranted. Therefore, the Wilde et al. study can not be taken as an adequate test of this issue. The interpretation of recall-recognition discrepancies as retrieval deficits should continue to be regarded as very plausible. What stands out as a robust result of the Wilde et al. study is the significantly greater heterogeneity of the Nondiscrepancy group as compared with the Discrepancy group, although this may well be a selection artifact, given a similar difference in the grouping variables. Regardless of whether this is the case, it underscores the necessity of greater attention to distributional problems, either by applying more stringent selection criteria or by using nonparametric statistical methods.

Cognition↗

PubMed2006

Adaptive Markov modeling for mutual-information-based, unsupervised MRI brain-tissue classification.

This paper presents a novel method for brain-tissue classification in magnetic resonance (MR) images that relies on a very general, adaptive statistical model of image neighborhoods. The method models MR-tissue intensities as derived from stationary random fields. It models the associated Markov statistics nonparametrically via a data-driven strategy. This paper describes the essential theoretical aspects underpinning adaptive, nonparametric Markov modeling and the theory behind the consistency of such a model. This general formulation enables the method to easily adapt to various kinds of MR images and the associated acquisition artifacts. It implicitly accounts for the intensity nonuniformity and performs reasonably well on T1-weighted MR data without nonuniformity correction. The method minimizes an information-theoretic metric on the probability density functions associated with image neighborhoods to produce an optimal classification. It automatically tunes its important internal parameters based on the information content of the data. Combined with an atlas-based initialization, it is completely automatic. Experiments on real, simulated, and multimodal data demonstrate the advantages of the method over the current state-of-the-art.

Algorithms↗

PubMed2001

Best harmony, unified RPCL and automated model selection for unsupervised and supervised learning on Gaussian mixtures, three-layer nets and ME-RBF-SVM models.

After introducing the fundamentals of BYY system and harmony learning, which has been developed in past several years as a unified statistical framework for parameter learning, regularization and model selection, we systematically discuss this BYY harmony learning on systems with discrete inner-representations. First, we shown that one special case leads to unsupervised learning on Gaussian mixture. We show how harmony learning not only leads us to the EM algorithm for maximum likelihood (ML) learning and the corresponding extended KMEAN algorithms for Mahalanobis clustering with criteria for selecting the number of Gaussians or clusters, but also provides us two new regularization techniques and a unified scheme that includes the previous rival penalized competitive learning (RPCL) as well as its various variants and extensions that performs model selection automatically during parameter learning. Moreover, as a by-product, we also get a new approach for determining a set of 'supporting vectors' for Parzen window density estimation. Second, we shown that other special cases lead to three typical supervised learning models with several new results. On three layer net, we get (i) a new regularized ML learning, (ii) a new criterion for selecting the number of hidden units, and (iii) a family of EM-like algorithms that combines harmony learning with new techniques of regularization. On the original and alternative models of mixture-of-expert (ME) as well as radial basis function (RBF) nets, we get not only a new type of criteria for selecting the number of experts or basis functions but also a new type of the EM-like algorithms that combines regularization techniques and RPCL learning for parameter learning with either least complexity nature on the original ME model or automated model selection on the alternative ME model and RBF nets. Moreover, all the results for the alternative ME model are also applied to other two popular nonparametric statistical approaches, namely kernel regression and supporting vector machine. Particularly, not only we get an easily implemented approach for determining the smoothing parameter in kernel regression, but also we get an alternative approach for deciding the set of supporting vectors in supporting vector machine.

Algorithms↗

PubMed1997

Predictors of intracranial pathologic findings in patients who seek emergency care because of headache.

BACKGROUND: Clinical criteria to select patients with headache in whom structural diagnostic studies (computed tomography) have a high yield disclosing intracranial pathologic findings, independent of abnormal findings on neurologic examination, have not been defined. OBJECTIVE: To determine which clinical characteristics predict the presence of intracranial pathologic findings, independently of neurologic examination, in patients with headache. DESIGN: Case-control, consecutive sample. SETTING: Major metropolitan trauma center emergency department. PATIENTS AND MATERIALS: Hospital records of 139 hospitalized and 329 randomly selected patients from 1720 nonhospitalized adult patients, consecutively evaluated for headache in the emergency department, were reviewed. Demographic data, clinical characteristics of the headache, results of neurologic and physical examinations, and diagnostic radiologic and laboratory results were correlated with final diagnosis and outcome at 6 months after emergency department visit. DATA ANALYSIS: Nonparametric statistical analysis. RESULTS: Intracranial pathologic findings were found in 18 (3.8%) of 468 patients. Acute onset and occipitonuchal location of headache, presence of associated symptoms, and patient age of 55 years or older were significantly associated with the finding of intracranial pathology, independently of the findings from neurologic examination. Abnormal findings on neurologic examination alone, whether focal or nonfocal, had a highly significant association and a positive predictive value for intracranial pathology of 39%. CONCLUSIONS: Abnormal results from neurologic examination are the best clinical parameters to predict structural intracranial pathology; however, in patients 55 years or older with headache of acute onset located in the occipitonuchal region that has associated symptoms, computed tomographic scan of the head is justified as part of their clinical evaluation independently of the findings of the neurologic examination.

Adult↗

PubMed2001

Functional decline in Parkinson disease.

OBJECTIVES: To determine the overall rate of functional decline and to assess the progression of different signs of Parkinson disease (PD). PATIENTS AND METHODS: Patients with clinically diagnosed PD followed up for at least 3 years were included in this study. Demographic and clinical data (including the Unified Parkinson's Disease Rating Scale [UPDRS]) were analyzed by the multivariate unbalanced repeated-measurements design using the mixed-effects model to study the association between different symptoms and various demographic variables. Regression models helped estimate the rates of progression of the disease in relationship to the various components of the UPDRS. Patients were categorized as having tremor-dominant or postural instability-gait difficulty-dominant PD and the 2 categories were compared for progression of their total UPDRS scores. DESIGN: A multivariate mixed-effects model was used to study the relationship between the different symptoms and various demographic variables. Nonparametric statistical tests were used to compare the progression of symptoms in the "on" (good function) state and the "off" (poor function) state groups for 2 age-at-onset categories (< or =57 and >57 years). RESULTS: Data from 1731 visits on 297 patients (181 men) followed up for an average of 6.36 years (range, 3-17 years) were analyzed. The annual rate of decline in the total UPDRS scores was 1.34 when assessed during the on state and 1.58 when assessed during the off state. Patients with an older age at onset had more rapid progression of PD than those with a younger age at onset. Furthermore, the older-onset group had statistically significantly more progression in mentation, freezing, and parts I (mentation) and II (activities of daily living) UPDRS subscores. Handwriting was the only component of the UPDRS score that did not notably deteriorate during the observation period. Regression analysis of 108 patients whose symptoms were rated during their off state showed a faster rate of cognitive decline as age at onset increased. The slopes (ie, the annual rates of decline) of progression in the UPDRS scores, when adjusted for age at the initial visit, were steeper for the postural instability--gait difficulty--dominant group compared with the tremor-dominant group. CONCLUSION: Based on longitudinal follow-up data, our findings provide evidence for a variable course of progression of the different PD symptoms, thus implying different biochemical or degenerative mechanisms for the various clinical features associated with PD.

Adolescent↗

PubMed2006

Predictors of visual outcome and choroidal neovascular membrane formation after traumatic choroidal rupture.

OBJECTIVE: To determine predictors of choroidal neovascularization (CNV) and visual outcome after traumatic choroidal rupture. METHODS: A retrospective review of patients with traumatic choroidal rupture diagnosed in the Retina Service, Massachusetts Eye and Ear Infirmary, Boston, between January 1993 and August 2001 was performed. Parametric and nonparametric statistical methods were used to evaluate visual prognosis, CNV, and retinal detachment after traumatic choroidal rupture. RESULTS: One hundred eleven cases were identified and reviewed. Visual acuity (VA) changes were recorded in all of the cases. Thirty-eight (34%) of the 111 patients recovered driving vision (VA > or =20/40). Rupture location was recorded in 107 cases. Recovery of driving vision was seen in 20 (59%) of 34 eyes with peripheral choroidal ruptures, 17 (22%) of 73 eyes with macular choroidal ruptures, 38 (38%) of 99 eyes without CNV, 1 (8%) of 12 eyes with CNV, 38 (40%) of 96 eyes without retinal detachment, and 1 (7%) of 15 eyes with retinal detachment. Older age and location of rupture within the arcades were positively associated with CNV formation (P = .04 and .03, respectively). Foveal location of rupture, multiple ruptures, and poor baseline VA were associated with failure to recover driving vision in univariate regression analyses. In multivariate analysis, rupture location and baseline VA were independently associated with visual outcome. Of 12 patients diagnosed with CNV, 5 were not treated, 4 were treated with argon laser photocoagulation, 1 was treated with surgery, 1 was treated with argon laser photocoagulation followed by surgery, and 1 was treated with verteporfin photodynamic therapy. CONCLUSIONS: Most patients with traumatic choroidal rupture do not achieve final VA of 20/40 or better. Poor visual outcome was most highly associated with a macular rupture and baseline VA of less than 20/40. The formation of CNV was most strongly associated with older age and macular choroidal rupture.

Adolescent↗

PubMed1982

Prolongation of the retinal mean circulation time in diabetes.

Retinal mean circulation time (MCT) and vascular sizes were measured in 21 normal individuals and 32 individuals with diabetes, and segmental blood flow (SBF) was calculated. The MCT was similar in the normal individuals (4.0 +/- 1.1 s) and the individuals with diabetes (4.2 +/- 1.9 s) when seven individuals with diabetes with prolonged but not quantifiable MCT were excluded. Including them by nonparametric statistics revealed that MCT was significantly longer in individuals with diabetes with proliferative retinopathy than in normal individuals or in individuals with diabetes with nonproliferative retinopathy. The prolonged MCT correlated significantly with advanced retinopathy as judged by leakage, neovascularization, and the need for photocoagulation therapy. Reduced SBF may account for the prolonged MCT, since the increase in vascular sizes observed failed to do so. However, pathologic vascular changes may alter the relation between SBF and MCT. Irrespective of implications about SBF, prolonged MCT, which indicates marked circulatory disturbance, represents an important new observation in diabetic retinopathy.

Adult↗

PubMed2005

Correlation between microtia and temporal bone malformation evaluated using grading systems.

OBJECTIVE: To evaluate the relationships between temporal bone abnormalities and the severity of microtia in Japanese patients using objective grading systems. DESIGN: Retrospective case series study conducted between 1992 and 2003. SETTING: Academic, tertiary care, referral medical center. PATIENTS: One hundred forty-two ears of 109 Japanese patients (85 male and 24 female patients; mean age, 12.8 years [range, 2-36 years]) with microtia. MAIN OUTCOME MEASURES: The severity of microtia was classified according to Marx classification. Developmental abnormalities of the temporal bone were evaluated by a computed tomographic (CT) scoring system modified after the system used by Jahrsdoerfer and colleagues, using high-resolution CT scans of the temporal bone. Correlations between the scores obtained from these 2 grading systems were evaluated using a nonparametric statistical method. RESULTS: Male preponderance and incidence of bilateral cases of approximately 30% were observed in our Japanese patients with microtia. There was no significant difference in the severity of microtia between unilateral and bilateral cases. The mean +/- SEM total points in the CT scoring system (full marks, 10) was 7.9 +/- 0.4 for grade I microtia, 6.6 +/- 0.6 for grade II, and 6.4 +/- 0.3 for grade III; the total points correlated inversely with the microtia grade. Development of the auricle correlated significantly with aeration in the middle ear spaces but not with ossicular development or formation of the oval/round windows. Proportion of acceptable surgical candidates according to the CT scoring system (>5 points) was 79% for grade I microtia, 52% for grade II microtia, and 65% for grade III microtia. CONCLUSION: The principle "the better developed the auricle, the better developed middle ear" was confirmed in Japanese patients with microtia; however, even with grade II/III microtia, more than half of the patients were considered suitable for atresia surgery.

Abnormalities, Multiple↗

PubMed1998

Decreased brain GABA(A)-benzodiazepine receptor binding in panic disorder: preliminary results from a quantitative PET study.

BACKGROUND: Positron emission tomography (PET) allows the measurement of benzodiazepine-gamma-aminobutyric acidA (GABA(A)) receptor kinetics. We employed flumazenil radiolabeled with carbon 11, a radioligand that labels the benzodiazepine site on the GABA(A) receptor, and fully quantitative, high-sensitivity PET to test the hypothesis that central benzodiazepine site binding is decreased in medication-free patients with panic disorder. METHODS: We compared 7 patients with panic disorder who had been off medication for at least 6 months and who had never abused alcohol with 8 healthy controls. The resulting parametric voxel-by-voxel maps were analyzed by voxel-based and region of interest-based methods using both parametric and nonparametric statistics. RESULTS: The major finding was that there is a global reduction in benzodiazepine site binding throughout the brain in patients with panic disorder compared with controls. There were sex differences in the 2 samples, but a separate analysis excluding women led to the same conclusions. In addition, the loci with the largest regional decrease in binding (right orbitofrontal cortex and right insula) were areas thought to be essential in the central mediation of anxiety. CONCLUSION: These results must be considered preliminary but are congruous with previous clinical psychopharmacologic evidence of involvement of the benzodiazepine-GABA(A) receptor and demonstrate that decreased flumazenil binding at this site may underlie panic disorder.

Adult↗

PubMed1996

Development of the orangutan permanent dentition: assessing patterns and variation in tooth development.

This study examines dental formation and alveolar emergence in a large cross-sectional sample composed primarily of wild-reared orangutans (N = 89) in order to provide information on the development of the permanent dentition in this hominoid and to address questions of variation in individual tooth formation, between teeth and between individuals. All specimens have been radiographed in lateral aspect and stages of crown and root formation recorded for all teeth. The ranges of crown and root formation of I1(1), C1(1), P4(4), M2(2), and M3(3) have been calculated relative to the stage of M1(1) development within a specific tooth quadrant. Then, for each specimen, BMDP scatterplot and nonparametric statistics have been used to graph changes in stages of these teeth relative to M1(1) stages and to examine relationships between pairs of upper and lower dental counterparts and between teeth of each jaw. Results indicate 1) high correlations between upper and lower tooth pairs and between many of the permanent teeth within individuals, 2) a relatively large range of variability in individual tooth development (multistage ranges relative to M1(1)), 3) greater variation in root development at emergence than earlier reports, and 4) evidence of variability within the sequence emergence pattern of the orangutan.

Animals↗