SEARCH · PubMed Health

Results for “Linear Models”

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 1,315 records · Page 73Linked to original sources

PubMedDate not supplied

Functional identification of the input-output transforms of mammalian motoneurones.

We studied the responses of rat hypoglossal and cat lumbar motoneurones to a variety of excitatory and inhibitory injected current transients during repetitive discharge. The amplitudes and time courses of the transients were comparable to those of the synaptic currents underlying postsynaptic potentials (PSPs) recorded in these cells. Poisson trains of these current transients were combined with an additional independent, high frequency random waveform to approximate band-limited white noise. The composite, white noise waveform was then superimposed on long duration suprathreshold current steps. We used the responses of the motoneurones to the white noise stimulus to derive zero-, first- and second-order Wiener kernels, which provide a quantitative description of the relation between injected current and discharge probability. The convolution integral computed for an injected current waveform and the first-order Wiener kernel provides the best linear prediction of the associated peristimulus time histogram (PSTH). This linear model provided good matches to most of the PSTHs compiled between the times of occurrence of individual current transients and motoneurone discharges. However, for the largest amplitude current transients, a significant improvement in the PSTH match was often achieved by expanding the model to include the convolution of the second-order Wiener kernel with the input. The overall transformation of current inputs into firing rate could be approximated by a second-order Wiener Model, i.e., a cascade of a dynamic, linear filter followed by a static non-linearity. At a given mean firing rate, the non-linear component of the motoneurone's response could be described by the square of the linear component multiplied by a constant coefficient. The amplitude of the response of the linear component increased with the average firing rate, whereas the value of the multiplicative coefficient in the nonlinear component decreased. As a result, the overall transform could be predicted from the mean firing rate and the linear impulse response, yielding a relatively simple, general description of the motoneurone's input-output function.

Action Potentials↗

PubMed2004

Contributions of Portuguese cattle breeds to genetic diversity using marker-estimated kinships.

The quantitative assessment of genetic diversity within and between populations is important for decision-making in genetic conservation plans. In our study, we applied the livestock core set method to define the contribution of 15 cattle breeds, 11 of which are Portuguese indigenous cattle breeds, to genetic diversity. In livestock core set theory genetic diversity is defined as the maximum genetic variance that can be obtained in a random-mating population that is bred from the populations present in that core set. Two methods to estimate marker-estimated kinships to obtain the contributions to the core set were used in this study: the weighted log-linear model (WLM) and the weighted log-linear mixed model (WLMM). The breeds that contributed most to diversity in the core set were Holstein-Friesian followed by the Portuguese Mertolenga and Cachena for both WLM and WLMM methods. The ranking of relative contributions of cattle breeds was maintained when we considered only the Portuguese cattle breeds. Furthermore, we were able to identify the marginal contributions and respective losses of diversity for each of the 11 Portuguese cattle breeds when we considered a subset of populations that are not threatened of being lost (the Safe set composed of the four exotic breeds present in this study). When WLM was used losses in genetic diversity ranged from 2.68 to 0.65% while the loss in founder genome equivalents ranged from 37.37 to 8.43% for Mertolenga and Brava de Lide breeds respectively. When WLMM was used losses in genetic diversity and founder genome equivalents were less extreme than for the WLM method, ranging from 1.27 to 0.69 and 26.8 to 12.99 respectively.

Animals↗

PubMed1999

Family and peer correlates of behavioral self-regulation in boys at risk for substance abuse.

Behavioral self-regulation (BSR), defined herein as the degree to which one can control one's own activity and reactivity to environmental stimuli, has been posited to be salient to the onset of adolescent substance abuse. The goal of this study was to clarify particular family and peer correlates of BSR in at-risk sons. Subjects were 10-through 12-year-old sons of substance-abusing fathers (high-average risk [HAR]; n = 176) and normal controls (low-average risk [LAR]; n = 199). A BSR latent trait was developed using multiple measures and multiple informants. Analyses included separate hierarchical linear regressions for HAR and LAR groups. In the hierarchical linear model for HAR sons, family dysfunction and deviant peer affiliation were significantly associated with BSR, whereas for LAR sons, only peer affiliation was significantly associated with BSR. The above family and peer correlates differed in proportions of variance explained for BSR in HAR and LAR sons. These findings extend previous studies by showing that, in a hierarchical linear model, BSR in HAR sons is associated with specific interpersonal, family, and peer factors. These findings suggest that empirical, theory-guided interventions to prevent worsening of BSR in HAR boys should address specific interpersonal, family, and peer factors.

Adult↗

PubMed2003

Managerial and environmental factors in the continuity of mental health care across institutions.

OBJECTIVE: The authors examined the association of continuity of care with factors assumed to be under the control of health care administrators and environmental factors not under managerial control. METHODS: The authors used a facility-level administrative data set for 139 Department of Veterans Affairs medical centers over a six-year period and supplemental data on environmental factors to conduct two types of analysis. First, simple correlations were used to examine bivariate associations between eight continuity-of-care measures and nine measures of the institutional environment and the social context. Second, to control for potential autocorrelation, multivariate hierarchical linear models with all nine independent measures were created. RESULTS: The strongest predictors of continuity of care were per capita outpatient expenditure and the degree of emphasis on outpatient care as measured by the percentage of all mental health expenditures devoted to outpatient care. The former was significantly associated with greater continuity of care on six of eight measures and the latter on seven of eight measures. The environmental factor of social capital (the degree of civic involvement and trust at the state level) was associated with greater continuity of care on five measures. The degree to which non-VA mental health services were funded in a state was unexpectedly found to be positively associated with greater continuity of care. In multivariate analysis using hierarchical linear modeling, significant relationships with continuity of care remained for per capita outpatient expenditures, overall outpatient emphasis, and social capital, but not for non-VA mental health funding. A linear term representing the year was positively and significantly associated with six of the eight examined continuity-of-care measures, indicating improvement in continuity of care for the period under study, although the explanation for this trend over time is unclear. CONCLUSIONS: Several factors potentially under managerial control are associated with increased mental health continuity of care.

Analysis of Variance↗

PubMed1990

Congenital malformations, stillbirths, and early mortality among the children of atomic bomb survivors: a reanalysis.

Of all the data sets pertinent to the estimation of the genetic risks to humans following exposure to ionizing radiation, potentially the most informative is that composed of the cohort of children born to atomic bomb survivors. We present here an analysis of the relationship between parental exposure history and untoward pregnancy outcomes within this cohort, using to the fullest extent possible the recently revised estimates of the doses received by their parents, the so-called DS86 doses. Available for study are 70,073 terminations, but DS86 doses have not been or presently cannot be computed on the parents of 14,770. The frequency of untoward pregnancy outcomes, defined as a pregnancy terminating in a child with a major congenital malformation, and/or stillborn, and/or dying in the first 14 days of life, increases with combined (summed) parental dose, albeit not significantly so. Under a standard linear model, when the sample of observations is restricted to those children whose parents have been assigned the newly established DS86 doses (n = 55,303), ignoring concomitant sources of variation and assuming a neutron RBE of 20, the estimated increase per sievert in the predicted frequency of untoward outcomes is 0.00354 (+/- 0.00343). After adjustment for concomitant sources of variation, the estimated increase per sievert in the proportion of such births is 0.00422 (+/- 0.00342) if the neutron RBE is assumed to be 20. A "one-hit" model with appropriate adjustments for extraneous sources of variation results in an almost identical value, namely, 0.00412 (+/- 0.00364). When the sample is extended to include parents lacking the full array of dose parameters necessary to calculate the DS86 dose, but sufficient for an empirical conversion of the previously employed T65DR dose system to its DS86 equivalent, we find under the linear model that the estimated increase per sievert in untoward pregnancy outcomes is some 31% higher than that published previously, 0.00264 (+/- 0.00277), assuming an RBE of 20, after adjustment for extraneous sources of variation. (Since a dose could not be calculated in 367 of the 70,073 outcomes, the n = 69,706). The corresponding value with the one-hit model is 0.00262 (+/- 0.00294).

Cohort Studies↗

PubMed1994

Extending the trend vector: the trend matrix and sample-based partial least squares.

Trend vector analysis [Carhart, R.E. et al., J. Chem. Inf. Comput. Sci., 25 (1985) 64], in combination with topological descriptors such as atom pairs, has proved useful in drug discovery for ranking large collections of chemical compounds in order of predicted biological activity. The compounds with the highest predicted activities, upon being tested, often show a several-fold increase in the fraction of active compounds relative to a randomly selected set. A trend vector is simply the one-dimensional array of correlations between the biological activity of interest and a set of properties or 'descriptors' of compounds in a training set. This paper examines two methods for generalizing the trend vector to improve the predicted rank order. The trend matrix method finds the correlations between the residuals and the simultaneous occurrence of descriptors, which are stored in a two-dimensional analog of the trend vector. The SAMPLS method derives a linear model by partial least squares (PLS), using the 'sample-based' formulation of PLS [Bush, B.L. and Nachbar, R.B., J. Comput.-Aided Mol. Design, 7 (1993) 587] for efficiency in treating the large number of descriptors. PLS accumulates a predictive model as a sum of linear components. Expressed as a vector of prediction coefficients on properties, the first PLS component is proportional to the trend vector. Subsequent components adjust the model toward full least squares. For both methods the residuals decrease, while the risk of overfitting the training set increases. We therefore also describe statistical checks to prevent overfitting. These methods are applied to two data sets, a small homologous series of disubstituted piperidines, tested on the dopamine receptor, and a large set of diverse chemical structures, some of which are active at the muscarinic receptor. Each data set is split into a training set and a test set, and the activities in the test set are predicted from a fit on the training set. Both the trend matrix and the SAMPLS approach improve the predictions over the simple trend vector. The SAMPLS approach is superior to the trend matrix in that it requires much less storage and CPU time. It also provides a useful set of axes for visualizing properties of the compounds. We describe a randomization method to determine the optimum number of PLS components that is very much faster for large training sets than leave-one-out cross-validation.

Cholinergic Agents↗

PubMedDate not supplied

Research on diabetes management and the family: a critique.

Research on diabetes management and the family has been traditionally viewed within a linear model, in which parental attitudes toward diabetes are seen as the principal influence on the child's adjustment and metabolic control. Recently the focus of research has shifted to the broader family milieu, with an emphasis on patterns of cooperation and conflict among all family members in implementing the treatment regimen. As investigators have begun to study the entire family, the linear model of parental influences has been overshadowed by a systems model of family interaction, based on the concept of mutual influences among all individuals in the family. Several methodological problems have characterized research in this area, such as inadequate assessments of family functioning, unreliable indices of metabolic control, and insensitivity to differences in age and disease variables. Future studies of diabetes management will have much to gain from consideration of the role of the father and siblings in treatment, attention to the diabetic child's impact on family functioning, and recognition of sources of support and stress outside the family that affect adaptation to diabetes.

Adaptation, Psychological↗

PubMed2005

Covariate detection in population pharmacokinetics using partially linear mixed effects models.

PURPOSE: To introduce partially linear mixed effects models (PLMEMs), to illustrate their use, and to compare the power and Type I error rate in detecting a covariate effect with nonlinear mixed effects modeling using NONMEM. METHODS: Sparse concentration-time data from males and females (1:1) were simulated under a 1-compartment oral model where clearance was sex-dependent. All possible combinations of number of subjects (50, 75, 100, 150, 250), samples per subject (2, 4, 6), and clearance multipliers (1 to 1.25) were generated. Data were analyzed with and without sex as a covariate using PLMEM (maximum likelihood estimation) and NONMEM (first-order conditional estimation). Four covariate screening methods were examined: NONMEM using the likelihood ratio test (LRT), PLMEM using the LRT, PLMEM using Wald's test, and analysis of variance (ANOVA) of the empirical Bayes estimates (EBEs) for CL treating sex as a categorical variable. The percent of simulations rejecting the null hypothesis of no covariate effect at the 0.05 level was determined. 300 simulations were done to calculate power curves and 1000 simulations were done (with no covariate effect) to calculate Type I error rate. Actual implementation of PLMEMs is illustrated using previously published teicoplanin data. RESULTS: Type I error rates were similar between PLMEM and NONMEM using the LRT, but were inflated (as high as 36%) based on PLMEM using Wald's test. Type I error rate tended to increase as the number of observations per subject increased for the LRT methods. Power curves were similar between the PLMEM and NONMEM LRT methods and were slightly more than the power curve using ANOVA on the EBEs of CL. 80% power was achieved with 4 samples per subject and 50 subjects total when the effect size was approximately 1.07, 1.07, 1.08, and 1.05 for LRT using PLMEMs, LRT using NONMEM, ANOVA on the EBEs, and Wald's test using PLMEMs, respectively. CONCLUSIONS: PLMEM and NONMEM covariate screening using the LRT had similar Type I error rates and power under the data generating model. PLMEMs offers a viable alternative to NONMEM-based covariate screening.

Computer Simulation↗

PubMed2005

Efficient multipole model and linear scaling of NDDO-based methods.

Fast growth of computational costs with that of the system's size is a bottleneck for the applications of traditional methods of quantum chemistry to polyatomic molecular systems. This problem is addressed by the development of linear (or almost linear) scaling methods. In the semiempirical domain, it is typically achieved by a series of approximations to the self-consistent field (SCF) solution. By contrast, we propose a route to linear scalability by modifying the trial wave function itself. Our approach is based on variationally determined strictly local one-electron states and a geminal representation of chemical bonds and lone pairs. A serious obstacle previously faced on this route were the numerous transformations of the two-center repulsion integrals characteristic for the neglect of diatomic differential overlap (NDDO) methods. We pass it by replacing the fictitious charge configurations usual for the NDDO scheme by atomic multipoles interacting through semiempirical potentials. It ensures invariance of these integrals and improves the computational efficiency of the whole method. We discuss possible schemes for evaluating the integrals as well as their numerical values. The method proposed is implemented for the most popular modified neglect of diatomic overlap (MNDO), Austin model 1 (AM1), and PM3 parametrization schemes of the NDDO family. Our calculations involving well-justified cutoff procedures for molecular interactions unequivocally show that the proposed scheme provides almost linear scaling of computational costs with the system's size. The numerical results on molecular properties certify that our method is superior with respect to its SCF-based ancestors.

Journal Article↗

PubMed1992

Time optimality in the control of human movements.

In a simulation study the control of maximally fast goal directed movements has been analyzed. For a simple linear model it is shown that the presence of a third input block reduces the movement duration. The time optimal size of the third block depends on the ratio of a neuromuscular time constant (first-order lag) and movement time. As a second step a non-linear muscle model was simulated. By an optimization of input parameters it was found that the time optimal input, as expected, switches between maximal agonist and maximal antagonist activation. As for the linear model, a third phase was required for an optimal movement. It was found that the third phase serves to compensate the slowly decaying antagonist force. Also an input similar to experimentally found activation patterns was simulated. This input contains a silent period between the first two bursts and the second and the third burst have submaximal amplitudes. This input led to a near time optimal movement with a duration 9% larger than the minimal duration but with largely reduced muscle forces. This suggests that a criterion is minimized which also takes into account the effort spent. Including gravity in the model indicates optimality of a silent period between the third phase and a final agonist activity to resist gravity. When assuming different dynamics for agonist and antagonist, the optimal switch times for agonist and antagonist no longer coincide, also after the three block pattern some extra activity is required to obtain a cancellation of the slowly decaying force in agonist and antagonist.

Biological Clocks↗

PubMed1994

Mortality from respiratory cancers (including lung cancer) among workers employed in formaldehyde industries.

A joint study on effects of formaldehyde exposure in industrial populations by the National Cancer Institute and the Formaldehyde Institute, Inc. (Blair et al. [1986]: J Natl Cancer Inst 76: 1071-1084; Blair and Stewart [1989]: J Occup Med 31: 881, Blair et al. [1990]: Am J Ind Med 17:683-700) reported no significant elevation in risk ratios standardized to the general population. Using the same data as Blair et al., we compared more exposed to less exposed workers to compute relative risk for respiratory and lung cancers using a multivariate, log-linear model incorporating factors for job type (hourly vs. salaried), cumulative exposure (0.1-0.5, 0.5-2, 2+ vs. < 0.1 ppm/years), length of exposure (1-5, 5-10, 10+ vs. < 1 years), and age. Models were fit for all workers, all males, all workers less than 65 years of age, and for all males less than 65 years of age. Hourly workers have a significantly elevated relative risk when compared to salaried workers. While only high levels of cumulative exposure show a significant elevation in relative lung cancer risk, trend analyses of the coefficients of a log-linear model show a significant trend of increasing risk with increasing formaldehyde exposure. The significantly elevated respiratory and lung cancer risk for workers younger than 65 may indicate a shift of respiratory cancer mortality towards younger ages among those occupationally exposed to formaldehyde.(ABSTRACT TRUNCATED AT 250 WORDS)

Adult↗

PubMed2002

Pharmacokinetics, tissue distribution, and metabolism of 17-(dimethylaminoethylamino)-17-demethoxygeldanamycin (NSC 707545) in CD2F1 mice and Fischer 344 rats.

PURPOSE: 17-(Dimethylaminoethylamino)-17-demethoxygeldanamycin (17DMAG) is an analogue of the benzoquinone ansamycin compound 17-(allylamino)-17-demethoxygeldanamycin (17AAG), which is currently being evaluated in clinical trials. Studies were performed in mice and rats to: (1) define the plasma pharmacokinetics, tissue distribution, and urinary excretion of 17DMAG after i.v. delivery; (2) define the bioavailability of 17DMAG after i.p. and oral delivery; (3) characterize the biliary excretion of 17DMAG after i.v. delivery to rats; and (4) characterize, if possible, any metabolites of 17DMAG observed in plasma, tissue, urine, or bile. MATERIALS AND METHODS: Studies were performed in female, CD2F1 mice or male Fischer 344 rats. In preliminary toxicity studies and subsequent i.v. pharmacokinetic studies in mice, 17DMAG i.v. bolus doses of 33.3, 50, and 75 mg/kg were used. In bioavailability studies, i.p. and oral 17DMAG doses of 75 mg/kg were used. In preliminary toxicity studies in rats, i.v. bolus doses of 10 and 20 mg/kg were used, and in i.v. pharmacokinetic studies 10 mg/kg was used. Compartmental and noncompartmental analyses were applied to the plasma concentration versus time data. In mice and rats, concentrations of 17DMAG were determined in multiple tissues. Urine was collected from mice and rats treated with each of the i.v. doses of 17DMAG mentioned above, and drug excretion was calculated until 24 h after treatment. Biliary excretion of 17DMAG and metabolites was studied in bile duct-cannulated rats given a 10 mg/kg i.v. bolus dose of 17DMAG. 17DMAG metabolites were identified with LC/MS. RESULTS: A 75 mg/kg dose of 17DMAG caused no changes in appearance, appetite, waste elimination, or survival of treated mice as compared to vehicle-treated controls. Bolus i.v. delivery of 17DMAG at 75 mg/kg produced "peak" plasma 17DMAG concentrations between 18 and 24.2 microg/ml in mice killed at 5 min after injection. Sequential reduction in the 17DMAG dose to 50 and 33.3 mg/kg resulted in "peak" plasma 17DMAG concentrations between 9.4 and 14.4, and 8.4 and 10.5 microg/ml, respectively. Plasma 17DMAG AUC increased from 362 to 674 and 1150 microg/ml x min when the 17DMAG dose increased from 33.3 to 50 and 75 mg/kg, respectively, corresponding to a decrease in 17DMAG CLtb from 92 ml/min per kg to 75 and 65 ml/min per kg. Plasma 17DMAG concentration versus time data were best fit by a two-compartment open linear model. No potential 17DMAG metabolites were observed in plasma. 17DMAG bioavailability was 100% and 50% after i.p. and oral delivery, respectively. In rats, an i.v. bolus dose of 10 mg/kg produced peak plasma 17DMAG concentrations between 0.88 and 1.74 microg/ml. Plasma 17DMAG concentrations had fallen below the lower limit of quantitation by 180 min and were best fit by a one-compartment open linear model. The plasma 17DMAG AUC was 104 microg/ml x min, corresponding to a 17DMAG CLtb of 96 ml/min per kg. 17DMAG distributed rapidly to all mouse and rat tissues except brain and testes. Only mouse liver contained materials consistent with potential metabolites of 17DMAG, but their concentrations were below the limit of quantitation of the HPLC assay used. Within the first 24 h after delivery, urinary excretion of 17DMAG by mice and rats accounted for 10.6-14.8% and 12.5-16%, respectively, of the delivered dose. By 15 min after i.v. delivery of 10 mg/kg of 17DMAG, rat bile contained 11 new materials with absorbance similar to that of 17DMAG. Four of these proposed metabolites had an Mr of 633, indicating addition of an oxygen. Two of these proposed metabolites had an Mr of 603, implying the loss of one methyl group, and one had an Mr of 589, implying the loss of two methyl groups. The remaining four proposed metabolites had an Mr of 566, 571, 629, and 645, respectively. Biliary excretion of 17DMAG and metabolites accounted for 4.7 +/- 1.4% of the delivered dose, with 17DMAG accounting for 50.7 +/- 3.4% of the biliary excretion. CONCLUSIONS: 17DMAG has excellent bioavailability when given i.p. and good bioavailability when given orally. 17DMAG is widely distributed to tissues and is quantitatively metabolized much less than is 17AAG. The pharmacokinetic and metabolite data generated should prove relevant to the design of additional preclinical studies as well as to contemplated clinical trials of 17DMAG and could be useful in their interpretation.

Animals↗

PubMed2005

The effect of the presence of others on caloric intake in homebound older adults.

BACKGROUND: Undernutrition in homebound older adults is a significant problem. The purpose of this study was to investigate the effect of the presence of others, both within the household and during meals, on caloric intake in homebound older adults. METHODS: In-depth interviews and three 24-hour dietary recalls were obtained from 50 older adults who were receiving home health services. Descriptive statistics were used to characterize participants, and hierarchical linear modeling was performed to evaluate predictors of caloric intake per meal. RESULTS: Participants' mean age was 77. Females composed 65% and African Americans composed 42% of the sample. Analyses are based on 553 meal observations. The majority (84%) of participants consumed all meals for each of the 3 days of data collection; however, they consumed an average of only 1305 calories per day. Hierarchical linear modeling analysis indicated that persons who had others present during meals consumed an average of 114.0 calories more per meal than those who ate alone (p =.009) and that women consumed 76.7 fewer calories per meal than did men (p =.045). The presence of others within the household had no effect on caloric intake. CONCLUSION: This research suggests that a simple and inexpensive way to increase caloric intake in homebound older adults is to make arrangements for family members or caregivers to eat with them.

Aged↗

PubMedDate not supplied

Completeness and validity of cancer registration in a major public referral hospital in Saudi Arabia.

BACKGROUND: The 1994 Saudi National Cancer Registry (NCR), a population-based registry, showed a crude incidence rate (CIR) of 39/100,000 for all cancers in the Saudi population. The low CIR suggested possible under-reporting, especially during the early years of operation. This study as aimed at estimating the number of missed cases due to under-reporting, and to assess the validity of reported data from a major public referral hospital in Riyadh city. MATERIALS AND METHODS: We compared cancer cases from the three data sources: medical records (MR), which were the original source of NCR cases; pathology reports (PR); and death certificates (DC). We estimated the missing cancer cases using the capture-recapture method with log-linear models of Fienberg to correct for interdependency between the three data sources. To assess the validity of the data, we reabstracted records of about 8% (39/4760 of previously reported cases from the same hospital. RESULTS: A total of 811 cancer cases were reported through the three sources, i.e., MR 611 (75.3%), PR 639 (78.8%) and DC 204 (25.2%). After fitting a series of log-linear models to the three sources of data, the three sources were found to be statistically dependent. Capture-repcapture method indicated that 384 cases were missed, giving an estimation of 1195 cancer cases to be reported. Using these 1195 estimated cases; the estimated ascertainment rates were 51% for medical records, 53% for pathology reports, 17% for death certificates, and 68% for the aggregated registry. In the validity assessment, major disagreement between the abstracted and reabstracted data was found to be highest for stage of disease (44%), followed by histology code and behavior (25.6%). Minor disagreements were most common for date of diagnosis (36%) and grade (36%). Overall, agreements were highest for laterality (95%), followed by primary site codes (90%) and basis of diagnosis (85%). Agreement of tumor description variables (site, histology, behavior, and stage) was 57%. CONCLUSION: Cancer registration will require substantial improvements in both completeness of reporting and data quality at the hospital level. Use of multiple data sources and estimation of missed cases will help ensure completeness of case registration.

Journal Article↗

PubMed1999

Application of Bayesian inference to fMRI data analysis.

The methods of Bayesian statistics are applied to the analysis of fMRI data. Three specific models are examined. The first is the familiar linear model with white Gaussian noise. In this section, the Jeffreys' Rule for noninformative prior distributions is stated and it is shown how the posterior distribution may be used to infer activation in individual pixels. Next, linear time-invariant (LTI) systems are introduced as an example of statistical models with nonlinear parameters. It is shown that the Bayesian approach can lead to quite complex bimodal distributions of the parameters when the specific case of a delta function response with a spatially varying delay is analyzed. Finally, a linear model with auto-regressive noise is discussed as an alternative to that with uncorrelated white Gaussian noise. The analysis isolates those pixels that have significant temporal correlation under the model. It is shown that the number of pixels that have a significantly large auto-regression parameter is dependent on the terms used to account for confounding effects.

Artifacts↗

PubMed2005

The cross-classified multilevel measurement model: an explanation and demonstration.

The link between the hierarchical generalized linear model (HGLM) and the Rasch model's parameterization has already been demonstrated by several researchers. Extensions have been described that include higher clustering levels to model more appropriately the contextual effects that are frequently encountered in educational research. However, pure hierarchies are relatively rare and instead cross-classified data structures are more frequently encountered. Cross-classified random effect modeling (CCREM) is still not commonly used. Use of CCREM in combination with the multilevel measurement model (MMM) has been recently introduced and is described further in the current study. Specifically, the link between the MMM and the CCREM MMM (termed "CCMMM" model) is provided. A dataset was simulated to demonstrate interpretation of the CCMMM model's parameters and to compare results under a CCMMM versus HGLM analysis. An Appendix is provided to demonstrate SAS GLIMMIX code used to estimate HGLM and CCMMM models' parameters.

Data Interpretation, Statistical↗

PubMed1992

Short-term projections of AIDS cases in Mexico.

We attempt to analyze and predict the behavior of the AIDS epidemic in Mexico. The reporting delay is corrected by using a cluster analysis, and the corrected data are used to make short-term projections by extrapolation, by fitting linear and log-linear models, and by back-calculation. The incubation period is assumed to have a Weibull distribution, and step functions are used for the infection functions. Most of the methods predict a mean of 25,000 accumulated cases by the end of 1993, and a comparison of the predictions with actual data up to November 1990 shows good agreement in all cases except the log-transformation linear model. The data for 1990 also show the reporting delay correction to be adequate in most cases.

Acquired Immunodeficiency Syndrome↗

PubMed1990

Chromatic mechanisms in striate cortex of macaque.

We measured the responses of 305 neurons in striate cortex to moving sinusoidal gratings modulated in chromaticity and luminance about a fixed white point. Stimuli were represented in a 3-dimensional color space defined by 2 chromatic axes and a third along which luminance varied. With rare exceptions the chromatic properties of cortical neurons were well described by a linear model in which the response of a cell is proportional to the sum (for complex cells, the rectified sum) of the signals from the 3 classes of cones. For each cell there is a vector passing through the white point along which modulation gives rise to a maximal response. The elevation (theta m) and azimuth (phi m) of this vector fully describe the chromatic properties of the cell. The linear model also describes neurons in l.g.n. (Derrington et al., 1984), so most neurons in striate cortex have the same chromatic selectivity as do neurons in l.g.n. However, the distributions of preferred vectors differed in cortex and l.g.n.: Most cortical neurons preferred modulation along vectors lying close to the achromatic axis and those showing overt chromatic opponency did not fall into the clearly defined chromatic groups seen in l.g.n. The neurons most responsive to chromatic modulation (found mainly in layers IVA, IVC beta, and VI) had poor orientation selectivity, and responded to chromatic modulation of a spatially uniform field at least as well as they did to any grating. We encountered neurons with band-pass spatial selectivity for chromatically modulated stimuli in layers II/III and VI. Most had complex receptive fields. Neurons in layer II/III did not fall into distinct groups according to their chromatic sensitivities, and the chromatic properties of neurons known to lie within regions rich in cytochrome oxidase appeared no different from those of neurons in the interstices. Six neurons, all of which resembled simple cells, showed unusually sharp chromatic selectivity.

Animals↗