SEARCH · PubMed Health

Results for “Multiple Classification Analysis”

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 685 records · Page 38Linked to original sources

PubMed2004

Long-term prognosis for childhood and juvenile absence epilepsy.

PURPOSE: To analyse prognostic factors for long term seizure remission in patients with childhood (CAE) and juvenile absence epilepsy (JAE). STUDY DESIGN: A retrospective analysis of a hospital based prevalence cohort. METHODS: The cohort consisted of 163 patients (104 females, 59 males) treated at the Universitatsklinik fur Neurologie, Innsbruck between 1970 and 1997. All had absences according to the ILAE classification. Follow up was in 1999 to 2000. We assessed multiple clinical and EEG factors as predictors of outcome and compared a classification according to the predominant pattern of seizure recurrence (pyknoleptic, PA or non pyknoleptic absence, NPA) with the ILAE classification with respect to prognosis. RESULTS: The mean age at seizure onset was 10.9 years (range, 3 to 27); age at follow up was 36.7 years (range, 13 to 81); duration of follow up was 25.8 years (range, 3 to 69). Sixty four patients (39 %) had CAE and 64 (39 %) JAE, while 35 (22%) had typical absences but could not be clearly defined as either CAE or JAE, and were therefore called "the overlap group". Patients with JAE or patients in the overlap group developed more often generalized tonic clonic seizures (GTCS) (p<0.001) and myoclonic attacks (p<0.05) during the course of the disease. At follow up 36 (56 %) of patients with CAE, 40 (62%) with JAE and 19 (54 %) of the overlap group were seizure free for at least two years (p=ns). When classified according to the predominant absence pattern at seizure onset 42 (51%) patients with PA and 53 (65%) with NPA were in remission (p=ns). In a stepwise binary logistic regression analysis the pattern of absence (PA or NPA) together with the later development of additional seizure types (myoclonias or GTCS), but not the CAE/JAE classification was predictive for long term lack of remission with a correct prediction of 66% of all patients. CONCLUSION: Only 58% of patients with absences were in remission after a long term follow up. CAE and JAE are closely related syndromes with large overlap of the age of onset. A classification according to the predominant seizure pattern at onset, together with later development of myoclonic attacks or GTCS is useful in predicting seizure remission in absence epilepsies.

Adolescent↗

PubMed2003

A multiple classifier system for early melanoma diagnosis.

Melanoma is the most dangerous skin cancer and early diagnosis is the key factor in its successful treatment. Well-trained dermatologists reach a diagnosis via visual inspection, and reach sensitivity and specificity levels of about 80%. Several computerised diagnostic systems were reported in the literature using different classification algorithms. In this paper, we will illustrate a novel approach by which a suitable combination of different classifiers is used in order to improve the diagnostic performances of single classifiers. We used three different kinds of classifiers, namely linear discriminant analysis (LDA), k-nearest neighbour (k-NN) and a decision tree, the inputs of which are 38 geometric and colorimetric features automatically extracted from digital images of skin lesions. Multiple classifiers were generated by combining the diagnostic outputs of single classifiers with appropriate voting schemata. This approach was evaluated on a set of 152 digital skin images. We compared the performances of multiple classifiers (2- and 3-classifier groups) between them and with respect to single ones (1-classifier group). We further compared the classifiers' performances with those of eight dermatologists. Classifiers' performances were measured in terms of distance from the ideal classifier. Compared with 1- and 2-classifier groups, performances of 3-classifier systems were significantly higher (P<0.0005 and P<0.001, respectively). No statistically significant differences were found between the 1- and 2-classifier groups (P=0.352). While the dermatologists group showed a level of performances significantly higher than the 1-classifier systems (P<0.020), no differences were found between the multiple classifier groups and the dermatologists groups, indicating comparable performances. This work suggests that a suitable combination of different kinds of classifiers can improve the performances of an automatic diagnostic system.

Algorithms↗

PubMed2004

Electrocardiographic poor R-wave progression: analysis of multiple criteria reveals little usefulness.

BACKGROUND: Poor or reverse R-wave progression (PRWP) is a common statement on electrocardiogram (ECG) interpretations, but its value in diagnosing anterior myocardial infarction (MI) is disputed. We assessed the accuracy of PRWP criteria in diagnosing anterior MI. METHODS: We searched MEDLINE (1960-1998) and found 3 criteria for PRWP. We included a modified version of the Marquette Muse system's criteria and multiple novel criteria. We interpreted resting ECGs of consecutive patients undergoing pharmacologic stress tests with dual isotope gated single photon emission computed tomography. Subjects with Q-wave anterior MI, bundle branch block, or Wolf-Parkinson-White syndrome were excluded. We established whether patients met the PRWP criteria. A nuclear cardiologist blinded to PRWP classifications reviewed the scintigrams. Chi2 methods were used for statistical analysis. RESULTS: Inclusion criteria were met by 122 subjects. The standard PRWP criteria were met in 15% to 42% of ECGs. Of subjects meeting PRWP criteria, 2% to 9% had anterior MI and 27% to 33% had anterior MI or ischemia. These proportions were similar to those expected by chance. The performance of PRWP criteria did not improve when subjects with electrocardiographic left ventricular hypertrophy were excluded or when more stringent criteria for right precordial R-wave amplitude were tested. CONCLUSIONS: In our study of patients undergoing cardiac stress tests, only a small percentage of patients who met various criteria for PRWP (a proportion no different than would be expected by chance) had anterior MI. Conclusions about the presence of anterior MI solely on the basis of PRWP have little usefulness.

Aged↗

PubMed2006

Continued exposure to silica after diagnosis of silicosis in Brazilian gold miners.

BACKGROUND: There is a paucity of studies analyzing the effect of continued silica exposure after the onset of silicosis with regard to disease progression. The present study investigates differences in clinical and radiological presentation of silicosis among former workers with a diagnosis of silicosis, and compares workers who continued to be exposed to silica with those who stopped silica exposure after having received their diagnosis. METHODS: A sample of 83 former gold miners with a median of 21 years from the first diagnoses of silicosis, had their clinical and occupational histories taken and underwent both chest radiography (International Labor Organization standards) and spirometry. Their silica exposure was assessed and an exposure index was created. The main outcome was the radiological severity of silicosis and tuberculosis (TB). The statistical analysis was done by multiple logistic regression. RESULTS: Among the 83 miners, 44 had continued exposed to silica after being diagnosed with silicosis. Continuation of silica exposure was associated with advanced radiological images of silicosis (X-ray classification in category 3, OR = 6.42, 95% CI = 1.20-34.27), presence of coalescence and/or large opacities (OR = 3.85, CI = 1.07-13.93), and TB (OR = 4.61, 95% CI = 1.14-18.71). CONCLUSIONS: Differential survival is unlikely to explain observed differences in silicosis progression. Results reinforce the recommendation that silica exposure should be halted at an early stage whenever X-ray is suggestive of the disease.

Adult↗

PubMed2005

Salt intake affects the relation between hypertension and the T-786C polymorphism in the endothelial nitric oxide synthase gene.

BACKGROUND: Recent genetic studies have shown an association between the T-786C polymorphism in the endothelial NO synthase gene and coronary artery diseases, but any possible association with hypertension has been controversial. Other studies indicate the effect of restricting salt intake differ depending on individual salt-sensitivity, and the mixture of different sensitivity in study subjects may obscure the results. The objective of this study was to investigate the gene-environment interaction between the salt intake and this polymorphism. METHODS: We genotyped 281 healthy men after excluding 37 men on hypertensive therapy (mean age 44.8 +/- 11.9 years) for the mutation, and evaluated their daily salt intake using a validated food frequency questionnaire. RESULTS: A quartile classification of salt intake revealed that the blood pressure of subjects with the mutation was significantly higher than that of subjects without the mutation, but only in the 4th quartile (the highest intake group). A multiple logistic regression analysis also showed that the presence of this mutation increased the risk of hypertension only in the 4th quartile (adjusted odds ratio = 6.38, P = .025). CONCLUSIONS: The presence of this mutation alone does not significantly increase the risk of hypertension. However, high salt intake interacts with the mutation and leads to a significant increase in the risk of hypertension. The T-786C mutation warrants being considered a candidate for further study with the aim of tailor-made hypertension prevention.

Adult↗

PubMed1995

Automated discrimination and quantification of idiopathic pulmonary fibrosis from normal lung parenchyma using generalized fractal dimensions in high-resolution computed tomography images.

RATIONALE AND OBJECTIVES: We computed generalized fractal dimensions for high-resolution computed tomography (HRCT) images to investigate their value in the discrimination and quantification of idiopathic pulmonary fibrosis (IPF) from normal lung parenchyma. METHODS: A probability distribution that was based on the pixel value in each image was used to compute capacity, information, and higher fractal dimensions for a series of 52 HRCT slices obtained from four patients. Qualitative classification of normal, mild, moderate, and severe IPF cases was achieved by computing the following parameter: DD = D0 - 2D1 + D2, where D0, D1, and D2 represents the capacity, information, and pair correlation dimensions, respectively. A multiple linear regression analysis using morphometric quantification for the set of 52 slices was tested for all possible combinations of the parameters D0, D1, D2, and D3. The generalizability of the model was tested by predicting the extent of IPF for each patient from a regression model computed with the remaining slices in the database. RESULTS: The best regression results were obtained using the independent parameters D1 and D2 to quantify the extent of diseased lung parenchyma. The technique was tested with 48 slices from 12 new patients. The results indicated that the extent of IPF could be predicted within the confidence limits given by the regression analysis. CONCLUSION: The extent of IPF can be predicted well within the 90% confidence interval given by the model. The width of the confidence interval decreases as the number of slices used in the linear regression model increases. This operator-independent quantitative technique may be useful in the follow-up of patients with IPF.

Confidence Intervals↗

PubMed1999

The effects of socioeconomic status and dental attendance on dental caries' experience, and treatment patterns in 5-year-old children.

OBJECTIVE: To compare the dental caries' experience and treatment received by 5-year-old children registered with a GDP. DESIGN: Retrospective case note review of all 5-year-old children registered with seven GDPs. SETTING: The study was carried out in 1996/7 in Wirral and North Cheshire in the north west of England. SUBJECTS AND MATERIALS: Clinical, demographic and attendance data were collected from each practice using a common data abstraction form. Subjects were categorised according to regular/irregular attenders, and into five groups ranging from affluent to deprived using the Super Profiles geodemographic classification. The relationships between disease experience, treatment, attendance and socioeconomic status were compared using cross-tabulations, t-tests and multiple linear regression. RESULTS: The dental records of 430 5-year-old children were available for analysis. Irregular attenders had significantly higher dmft, dt and mt, and fewer filled teeth. Only 29% of disease experience of regular attenders was treated by restoration. Both socioeconomic status and visiting behaviour exerted significant independent effects on dmft, but dental attendance alone had a significant effect on ft. CONCLUSIONS: Significant inequalities remain in the disease experience and service use of young children. Regularly attending children have less than a third of their diseased teeth restored. Consensus is needed across the profession on the care of the diseased deciduous dentition.

Child, Preschool↗

PubMed2003

Transmission mode time-reversal super-resolution imaging.

The theory of time-reversal super-resolution imaging of point targets embedded in a reciprocal background medium [A. J. Devaney, "Super-resolution imaging using time-reversal and MUSIC," J. Acoust. Soc. Am. (to be published)] is generalized to the case where the transmitter and receiver sensor arrays need not be coincident and for cases where the background medium can be nonreciprocal. The new theory developed herein is based on the singular value decomposition of the generalized multistatic data matrix of the sensor system rather than the standard eigenvector/eigenvalue decomposition of the time-reversal matrix as was employed in the above-mentioned work and other treatments of time-reversal imaging [Prada, Thomas, and Fink, "The iterative time reversal process: Analysis of the convergence," J. Acoust. Soc. Am. 97, 62 (1995); Prada et al., "Decomposition of the time reversal operator: Detection and selective focusing on two scatterers," J. Acoust. Soc. Am. 99, 2067 (1996)]. A generalized multiple signal classification (MUSIC) algorithm is derived that allows super-resolution imaging of both well-resolved and non-well-resolved point targets from arbitrary sensor array geometries. MUSIC exploits the orthogonal nature of the scatterer and noise subspaces defined by the singular vectors of the multistatic data matrix to form scatterer images. The time-reversal/MUSIC algorithm is tested and validated in two computer simulations of offset vertical seismic profiling where the sensor sources are aligned along the earth's surface and the receiver array is aligned along a subsurface borehole. All results demonstrate the high contrast, high resolution imaging capabilities of this new algorithm combination when compared with "classical" backpropagation or field focusing. Above and beyond the application of seismo-acoustic imaging, the time-reversal super-resolution theory has applications in ocean acoustics for target location, and ultrasonic nondestructive evaluation of parts.

Acoustics↗

PubMed1999

Screening of obstructive sleep apnea syndrome by heart rate variability analysis.

BACKGROUND: Enhanced nocturnal heart rate variability (HRV) has been evoked in sleep-related breathing disorders. However, its capacity to detect obstructive sleep apnea syndrome (OSAS) has not been systematically determined. Thus, we evaluated the discriminant power of HRV parameters in a first group of patients (G1) and validated their discriminant capacity in a second group (G2). METHODS AND RESULTS: In G1, 39 of 91 patients (42.8%) were identified as diseased by polysomnography, as were 24 of 52 patients (46%) in G2. Time-domain HRV variables (SD of NN intervals [SDNN], mean of the standard deviations of all NN intervals for all consecutive 5-minute segments of the recording [SDNN index], square root of the mean of the sum of the squares of differences between adjacent normal RR intervals [r-MSSD], and SD of the averages of NN intervals in all 5-minute segments of the recording [SDANN]) were calculated for daytime and nighttime periods, as well as the differences between daytime and nighttime values (Delta[D/N]). Correlations between HRV variables and OSAS status were analyzed in G1 by use of receiver-operating characteristic (ROC) curves and logistic regression analysis. By ROC curve analysis, 7 variables were significantly associated with OSAS. After adjustment for other variables through multiple logistic regression analysis, Delta[D/N]SDNN index and Delta[D/N] r-MSSD remained significant independent predictors of OSAS, with ORs of 8.22 (95% CI, 3.16 to 21.4) and 2.86 (95% CI, 1.21 to 6.75), respectively. The classification and regression tree methodology demonstrated a sensitivity reaching 89.7% (95% CI, 73.7 to 97.7) with Delta[D/N] SDNN index and a specificity of 98.1% (95% CI, 86.4 to 100) with Delta[D/N] SDNN using appropriate thresholds. These thresholds, applied to G2, yielded a sensitivity of 83% using Delta[D/N] SDNN index and a specificity of 96.5% using Delta[D/N] SDNN. CONCLUSIONS: Time-domain HRV analysis may represent an accurate and inexpensive screening tool in clinically suspected OSAS patients and may help focus resources on those at the highest risk.

Adult↗

PubMed1989

Evaluation of evoked potentials and cerebrospinal fluid analysis in the differential diagnosis of multiple sclerosis.

During the 6 years to December 1988, 191 patients underwent evoked potential (EP) studies and cerebrospinal fluid (CSF) analysis for oligoclonal bands at the Queen Elizabeth Hospital, for the differential diagnosis of multiple sclerosis (MS). Clinical data at the time of study separated patients into 3 groups, as follows: (i) multiple sclerosis (n = 90) - McDonald and Halliday Classification, 1977, (ii) other neurological disease (n = 82) and (iii) no neurological disease (n = 19). In cases of clinically definite MS, visual evoked potentials (VEPs) were abnormal and oligoclonal bands were detected in 64% and 59% of cases, respectively. However, only 15% of patients with suspected MS had abnormal VEPs, and only 23% had oligoclonal bands. Other studies have shown figures differing from these, though not necessarily significantly. We found a substantial number of EP (20%) but few CSF (4%) abnormalities in disorders other than MS, and no abnormalities in cases without neurological disease. The various figures for abnormal results in cases assessed for the differential diagnosis of MS are influenced not only by laboratory methods, but by the degree of clinical suspicion in relation to the cases selected, as well as by differences in populations from which cases are derived. Long-term prospective studies of diagnostically indeterminate cases are still required to determine the diagnostic weighting that can be applied on the basis of abnormal investigational results. Magnetic resonance imaging will not resolve these questions since it has limitations of its own, particularly with regard to specificity.

Adult↗

PubMed1999

Self-organizing map for the classification of normal and disordered female voices.

The goal of this research was to train a self-organizing map (SOM) on various acoustic measures (amplitude perturbation quotient, degree of voice breaks, rahmonic amplitude, soft phonation index, standard deviation of the fundamental frequency, and peak amplitude variation) of the sustained vowel /a/ to enhance visualization of the multidimensional nonlinear regularities inherent in the input data space. The SOM was trained using 30 spasmodic dysphonia exemplars, 30 pretreatment functional dysphonia exemplars, 30 post-treatment functional dysphonia exemplars, and 30 normal voice exemplars. After training, the classification performance of the SOM was evaluated. The results indicated that the SOM had better classification performance than that of a stepwise discriminant analysis over the original data. Analysis of the weight values across the SOM, by means of stepwise discriminant analysis, revealed the relative importance of the acoustic measures in classification of the various groups. The SOM provided both an easy way to visualize multidimensional data, and enhanced statistical predictability at distinguishing between the various groups (over that conducted on the original data set). We regard the results of this study as a promising initial step into the use of SOMs with multiple acoustic measures to assess phonatory function.

Adult↗

PubMed2006

Distinguishing HIV-1 drug resistance, accessory, and viral fitness mutations using conditional selection pressure analysis of treated versus untreated patient samples.

BACKGROUND: HIV can evolve drug resistance rapidly in response to new drug treatments, often through a combination of multiple mutations 123. It would be useful to develop automated analyses of HIV sequence polymorphism that are able to predict drug resistance mutations, and to distinguish different types of functional roles among such mutations, for example, those that directly cause drug resistance, versus those that play an accessory role. Detecting functional interactions between mutations is essential for this classification. We have adapted a well-known measure of evolutionary selection pressure (Ka/Ks) and developed a conditional Ka/Ks approach to detect important interactions. RESULTS: We have applied this analysis to four independent HIV protease sequencing datasets: 50,000 clinical samples sequenced by Specialty Laboratories, Inc.; 1800 samples from patients treated with protease inhibitors; 2600 samples from untreated patients; 400 samples from untreated African patients. We have identified 428 mutation interactions in Specialty dataset with statistical significance and we were able to distinguish primary vs. accessory mutations for many well-studied examples. Amino acid interactions identified by conditional Ka/Ks matched 80 of 92 pair wise interactions found by a completely independent study of HIV protease (p-value for this match is significant: 10-70). Furthermore, Ka/Ks selection pressure results were highly reproducible among these independent datasets, both qualitatively and quantitatively, suggesting that they are detecting real drug-resistance and viral fitness mutations in the wild HIV-1 population. CONCLUSION: Conditional Ka/Ks analysis can detect mutation interactions and distinguish primary vs. accessory mutations in HIV-1. Ka/Ks analysis of treated vs. untreated patient data can distinguish drug-resistance vs. viral fitness mutations. Verification of these results would require longitudinal studies. The result provides a valuable resource for AIDS research and will be available for open access upon publication at http://www.bioinformatics.ucla.edu/HIV.

Journal Article↗

PubMed2004

Microarray data analysis: from hypotheses to conclusions using gene expression data.

We review several commonly used methods for the design and analysis of microarray data. To begin with, some experimental design issues are addressed. Several approaches for pre-processing the data (filtering and normalization) before the statistical analysis stage are then discussed. A common first step in this type of analysis is gene selection based on statistical testing. Two approaches, permutation and model-based methods are explained and we emphasize the need to correct for multiple testing. Moreover, powerful approaches based on gene sets are mentioned. Clustering of either genes or samples is frequently performed when analyzing microarray data. We summarize the basics of both supervised and unsupervised clustering (classification). The latter may be of use for creating diagnostic arrays, for example. Construction of biological networks, such as pathways, is a statistically challenging but complex task that is a relatively new development and hence mentioned only briefly. We finish with some remarks on literature and software. The emphasis in this paper is on the philosophy behind several statistical issues and on a critical interpretation of microarray related analysis methods.

Cluster Analysis↗

PubMed1996

Spatiotemporal analysis of prepyriform, visual, auditory, and somesthetic surface EEGs in trained rabbits.

1. Spatial ensemble averages were computed for 64 traces of electroencephalograms (EEGs) simultaneously recorded from 8 x 8 arrays over the epidural surfaces of the prepyriform cortex (PPC) and visual, somatic, and auditory cortices. They revealed a common waveform across each array. Examination of the spatial amplitude modulation (AM) of the waveform revealed classifiable spatial pattern in short time segments. The AM patterns varied within trials after presentation of identical conditioned stimuli, and also between trials with differing stimuli. 2. PPC EEGs revealed strong correlates with the respiratory rhythm; neocortical EEGs did not. 3. Time ensemble averaging of the PPC EEG attenuated the oscillatory bursts, indicating that olfactory gamma oscillations (20-80 Hz) were not phase-locked to the times of stimulus delivery but instead to inhalations. Time ensemble averages of neocortical recordings across trials revealed average evoked potentials starting 30-50 ms after the arrival of the stimulus. 4. Average temporal fast Fourier transform (FFT) power spectral densities (PSDs) from pre- and poststimulus PPC EEG segments revealed a peak of gamma activity in olfactory bursts. 5. The logarithm of the average temporal FFT PSDs from pre- and poststimulus neocortical EEG segments, when plotted against log frequency, revealed 1/f-type spectra in both pre- and poststimulus segments for negative/aversive conditioned stimuli (CS-) and positive/rewarding conditioned stimuli (CS+). The alpha'- and beta'-coefficients from the regression of Eq. 2 onto the average PSDs were significantly different between pre- and poststimulus segments, owing to the evoked potentials, but not between CS- and CS+ stimulus segments. 6. Spatiotemporal patterns were invariant over all frequency bins in the 1/f domain (20-100 Hz). Spatiotemporal patterns in the 2- to 20-Hz domain progressively differed from the invariant patterns with decreasing frequency. 7. In the spatial frequency domain, the logarithm of the average spatial FFT power spectra from pre- and poststimulus neocortical EEG segments, when plotted against the log spatial frequency, fell monotonically from the maximum at the lowest spatial frequency, downwardly curving to a linear 1/f spectral domain. This curve in the 1/f spectral domain extended from 0.133 to 0.880 cycles/mm in the PPC and from 0.095 to 0.624 cycles/mm in the neocortices. 8. Methods of FFT and principal component analysis (PCA) EEG decomposition were used to extract the broad-spectrum waveform common to all 64 EEGs from an array. AM patterns for the FFT and PCA components were derived by regression. They were shown by cross-correlation to yield spatial patterns that were equivalent to each other and to AM patterns from calculation of the 64 root-mean-square amplitudes of the segments. 9. Each spatial AM pattern was expressed by a 1 x 64 column vector and a point in 64-space. Similar patterns formed clusters, and dissimilar patterns gave multiple clusters. A statistical test was devised to evaluate dissimilarity by a Euclidean distance metric in 64-space. 10. Significant spatial pattern classification of CS- versus CS+ trials (below the 1% confidence limit for 20 of each) was found in discrete temporal segments of poststimulus data after digital temporal and spatial filter optimization. 11. Varying the analysis window duration from 10 to 500 ms yielded a window length of 120 ms as optimal for pattern classification. A 120-ms window was subsequently stepped across each record in overlapping intervals of 20 ms. Windows in which episodic, significant CS+/CS- differences occurred lasted 50-200 ms and were separated by 100-200 ms in the poststimulus period. 12. Neocortical spatial patterns changed under reinforcement contingency reversal, showing a lack of invariance in respect to stimuli and a dependence on context and learning, as previously found for the olfactory bulb and PPC.

Animals↗

PubMed2004

Malnutrition and the severity of lung disease in adults with pulmonary tuberculosis in Malawi.

SETTING: Zomba Central Hospital, Zomba, Malawi. OBJECTIVE: To examine the relationship between malnutrition and the severity of lung disease in human immunodeficiency virus (HIV) positive and negative adults with pulmonary tuberculosis (PTB). DESIGN: Cross-sectional study. METHODS: Chest radiographs and anthropometric measurements were obtained and bioelectrical impedance analysis was conducted in sputum-positive patients with pulmonary tuberculosis. Lung disease in chest radiographs was graded as normal, minimal, moderately advanced and far advanced according to a conventional classification system. RESULTS: Among 319 adults with PTB with or without HIV co-infection, body mass index (BMI), fat mass and phase angle were independently associated with increasing severity of lung disease. Multiple logistic regression analyses showed that BMI, fat mass and phase angle were associated with increasing severity of lung disease among 236 HIV-positive adults, when adjusted for sex, age, and plasma HIV load. CONCLUSION: The severity of lung disease in adults with PTB is associated with the extent of malnutrition, as reflected by BMI and body composition studies using bioelectrical impedance analysis.

Adult↗

PubMed2003

Statistical method for analysis of the disease curve in animals with experimental autoimmune encephalomyelitis.

Experimental autoimmune encephalomyelitis (EAE) is a procedure used in the laboratory to examine drugs that may have utility in treating multiple sclerosis (MS). The problem of modeling the disease curve in animals with EAE is studied. The classification of animals after each experiment is considered and the chi-square test is proposed to test a homogeneity between treatment groups. A mixture type of nonlinear mixed-effects model with repeated measurements is considered, assuming that the onset and/or remission of disease is the fixed effect as well as the random effect. Statistical inference on the parameters of the disease curves is discussed. The proposed model is shown to be efficient for comparing the disease curves with different treatments. Examples concerning the study of the effects of test compounds in EAE are presented to illustrate the proposed model and statistical methodologies.

Animals↗

PubMed2004

Recursive partitioning analysis classifications I and II: applicability evaluated in a randomized trial for resected single brain metastases.

PURPOSE: Radiation Therapy Oncology Group (RTOG) recursive partitioning analysis (RPA) prognostic classes I and II for patients with brain metastases is derived from a database made up primarily of patients with unresected and multiple metastases. An analysis of a previously published randomized trial was performed to determine the applicability of these RPA prognostic classes in the setting of resection of single metastases to the brain. PATIENTS AND METHODS: Ninety-five patients with single metastases to the brain that were treated with complete surgical resection entered this study. Patients were randomly assigned to treatment with postoperative whole brain radiotherapy (WBRT) (n = 49 patients) or no further brain treatment (n = 46 patients). All patients entered on this study had a Karnofsky performance status of > or =70. Therefore, although the RTOG RPA has 3 classes, only patients with RPA classes I (n = 26) or II (n = 69) were eligible for this study analysis. RESULTS: For RPA class I, the median survival was 10.9 months versus 9.8 months for class II patients (P = 0.45). Multivariate analysis showed that only postoperative WBRT, independent of RPA class I or II, lessened the risk of brain tumor recurrence (P < 0.0001). CONCLUSION: This analysis of a randomized trial evaluating postoperative WBRT in the treatment of single metastases to the brain showed no difference in survival between RPA class I or II patients. In addition, the use of postoperative WBRT after complete surgical resection of single brain metastases results in substantially better control of disease in the brain independent of RPA classes I or II.

Adult↗

PubMed2005

Bayesian model averaging of Bayesian network classifiers over multiple node-orders: application to sparse datasets.

Bayesian model averaging (BMA) can resolve the overfitting problem by explicitly incorporating the model uncertainty into the analysis procedure. Hence, it can be used to improve the generalization performance of Bayesian network classifiers. Until now, BMA of Bayesian network classifiers has only been performed in some restricted forms, e.g., the model is averaged given a single node-order, because of its heavy computational burden. However, it can be hard to obtain a good node-order when the available training dataset is sparse. To alleviate this problem, we propose BMA of Bayesian network classifiers over several distinct node-orders obtained using the Markov chain Monte Carlo sampling technique. The proposed method was examined using two synthetic problems and four real-life datasets. First, we show that the proposed method is especially effective when the given dataset is very sparse. The classification accuracy of averaging over multiple node-orders was higher in most cases than that achieved using a single node-order in our experiments. We also present experimental results for test datasets with unobserved variables, where the quality of the averaged node-order is more important. Through these experiments, we show that the difference in classification performance between the cases of multiple node-orders and single node-order is related to the level of noise, confirming the relative benefit of averaging over multiple node-orders for incomplete data. We conclude that BMA of Bayesian network classifiers over multiple node-orders has an apparent advantage when the given dataset is sparse and noisy, despite the method's heavy computational cost.

Algorithms↗