Search PubMed⌕ Search

SEARCH · Search PubMed

Results for “Predictive”

Search indexed PubMed citations on genomics, clinical trials, systematic reviews and public health. Explore titles, authors and supplied subject terms, then open the PubMed record.

Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.

At least 1,045 records · Page 58Linked to original sources
PubMed2005

Prediction of in vivo drug-drug interactions from in vitro data: impact of incorporating parallel pathways of drug elimination and inhibitor absorption rate constant.

AIMS: Success of the quantitative prediction of drug-drug interactions via inhibition of CYP-mediated metabolism from the inhibitor concentration at the enzyme active site ([I]) and the in vitro inhibition constant (K(i)) is variable. The aim of this study was to examine the impact of the fraction of victim drug metabolized by a particular CYP (f(mCYP)) and the inhibitor absorption rate constant (k(a)) on prediction accuracy. METHODS: Drug-drug interaction studies involving inhibition of CYP2C9, CYP2D6 and CYP3A4 (n = 115) were investigated. Data on f(mCYP) for the probe substrates of each enzyme and k(a) values for the inhibitors were incorporated into in vivo predictions, alone or in combination, using either the maximum hepatic input or the average systemic plasma concentration as a surrogate for [I]. The success of prediction (AUC ratio predicted within twofold of in vivo value) was compared using nominal values of f(mCYP) = 1 and k(a) = 0.1 min(-1). RESULTS: The incorporation of f(mCYP) values into in vivo predictions using the hepatic input plasma concentration resulted in 84% of studies within twofold of in vivo value. The effect of k(a) values alone significantly reduced the number of over-predictions for CYP2D6 and CYP3A4; however, less precision was observed compared with the f(mCYP). The incorporation of both f(mCYP) and k(a) values resulted in 81% of studies within twofold of in vivo value. CONCLUSIONS: The incorporation of substrate and inhibitor-related information, namely f(mCYP) and k(a), markedly improved prediction of 115 interaction studies with CYP2C9, CYP2D6 and CYP3A4 in comparison with [I]/K(i) ratio alone.

Area Under Curve↗
PubMed2005

The predictive value of specific immunoglobulin E levels in serum for the outcome of oral food challenges.

BACKGROUND: Specific serum IgE is considered as one of the important diagnostic measures in the diagnostic work-up of food allergy. OBJECTIVE: To evaluate the role of specific serum IgE in predicting the outcome of oral food challenges, and to determine threshold concentrations of specific serum IgE that could render double-blind, placebo-controlled food challenges unnecessary. METHODS: In 501 children (median age 13 months), 992 controlled oral challenges were performed with cow's milk (CM), hen's egg (HE), wheat and soy. 440/501 (88%) children suffered from atopic dermatitis. For all children, specific IgE concentrations in serum were determined. Sensitivity, specificity, positive and negative predictive values, receiver operator characteristics-curves as well as predictive decision points were calculated. RESULTS: Four hundred and forty-five out of 992 oral food challenges with allergens were assessed as positive. Sensitivity of specific serum IgE was 97% for HE, 83% for CM, 69% for soy, and 79% for wheat. Specificity was 51% for HE, 53% for CM, 50% for soy, and 38% for wheat. Calculating 90%, 95% and 99% predicted probabilities using logistic regression revealed predictive decision points of 6.3, 12.6, and 59.2 kU/L for HE, respectively. Subdividing our children in those of below or above 1 year of age resulted in a markedly different predicted probability for HE. For CM, only the 90% predicted probability (88.8 kU/L) could be calculated. No decision points could be determined for CM, wheat and soy. CONCLUSION: In general, specific serum IgE levels showed a correlation with the outcome of positive oral food challenges for CM and HE. Meaningful predictive decision points can be calculated for HE, which may help to avoid oral food challenges in some cases. However, data need to be ascertained for each allergen separately. Furthermore, the age of the patient population under investigation must also be taken into account.

Adolescent↗
PubMed2004

Value of single and paired serum human chorionic gonadotropin measurements in predicting outcome of in vitro fertilisation pregnancy.

OBJECTIVES: To assess whether paired human chorionic gonadotropin (hCG) measurements in early pregnancy are more effective than a single measurement, in predicting the outcome for an in vitro fertilisation pregnancy. DESIGN: Retrospective analysis. SETTING: Westmead Fertility Centre, Westmead Hospital, Sydney, Australia. MATERIALS AND METHODS: Serial hCG measurements in 143 patients at Westmead Fertility Centre, from August 1997 to April 2000, were studied retrospectively. The predictive value of single hCG measurements relative to the published assay reference ranges were evaluated. The predictive value of serial hCG levels in predicting pregnancy outcome was assessed separately. Normal daily rate of rise (ROR) of hCG was defined as the mean ROR for ongoing pregnancies +/- 1 SD. Abnormal daily ROR was defined as a daily increase in hCG less than the mean ROR for ongoing pregnancies--1 SD. MAIN OUTCOME MEASURES: Viability of the pregnancy at 20 weeks' gestation. RESULTS: An initial hCG measurement below the 5th centile reference limit for gestation has 85% (confidence interval (CI) 75-92%) positive predictive value for non-viability, with a sensitivity of 40% (CI 33-48). Serial testing of borderline samples for ROR did not improve positive predictive value (70%: CI 50-86%) or sensitivity (30%: CI 20-43%) in identifying non-ongoing pregnancies. CONCLUSIONS: In assisted reproductive technologies pregnancies, comparison of a single hCG value with appropriate reference ranges enables approximately 40% of non-viable pregnancies to be identified with a high positive predictive value. Repeated measurements did not contribute further to the predictive value.

Adult↗
PubMed1993

Are activity-based assessments of balance and gait in the elderly predictive of risk of falling and/or type of fall?

OBJECTIVE: To determine whether an activity-based test of balance and gait is predictive of the risk of: (1) falling in situations that are related to specific tasks evaluated as part of the test, (2) experiencing falls precipitated by different classes of biomechanical events, or (3) falling in general; and to compare the predictive ability of the activity-based test for the falls described in (2) and (3) to that of a posturography test that has been found previously to be predictive of falling risk. DESIGN: Cohort study. SETTING: Baseline tests performed in balance laboratory; subsequent history of falling monitored prospectively for 1 year in two residential-care facilities. PARTICIPANTS: Seventeen male and 83 female consecutive volunteers (mean age = 83, SD = 6) who were independent in activities of daily living and able to stand unaided. MEASUREMENTS: Independent variables were derived from an activity-based balance-and-gait test and a posturography test. Dependent variables were the numbers of subjects with one or more: (1) falls in specific situations related to activity-based test items, (2) falls related to general classes of biomechanical precipitant, and (3) falls in general. MAIN RESULTS: Subjects who were rated as "abnormal" in activity-based test items related to transfers, turning or reaching were more likely to experience one or more falls in related situations in everyday life. Activity-based scores were predictive of risk of experiencing falls with no obvious biomechanical precipitant and falls precipitated by center-of-mass perturbation, but not falls precipitated by base-of-support perturbation. In comparison, a posturographic measure of spontaneous medial-lateral postural sway (blindfolded conditions) failed to predict falls having no biomechanical precipitant, but provided the best predictions of both center-of-mass and base-of-support falls, as well as risk of falling in general. CONCLUSIONS: Activity-based testing of certain tasks (transfer, turning, reaching) may be useful in indicating a specific need for intervention to reduce the risk of falling during related everyday activities. In terms of predicting falling risk, a static posturography test may provide better prediction overall of the different classes of falls and may be useful as a quick and simple screening tool to help identify high-risk individuals.

Accidental Falls↗
PubMed1997

The synergy of low lung function and low body mass index predicting all-cause mortality among older Japanese-American men.

OBJECTIVE: To assess the joint characteristics of low standardized weight and compromised pulmonary function in predicting all-cause mortality. DESIGN: A population-based, prospective cohort study. SETTING: Oahu Island, Hawaii. PARTICIPANTS: Surviving Japanese-American men of the Honolulu Heart Program cohort, 71 to 93 years of age (N = 3059). MEASUREMENTS: Body mass index (BMI-weight in kilograms/square of height in meters) and 1-second forced expiratory volume (FEV1) as a percentage of age- and height-predicted FEV1 from the 1991 to 1993 examination of the cohort. Mortality data derived from the ongoing tracking of deaths of the cohort. Relations of selected risk factors among joint levels of BMI (< or = 21, > 21 to < 25, > or = 25 kg/m2) and percent predicted FEV1 (< or = 70%, > 70%) were determined. The impact of these covariates on relations between joint BMI/percent predicted FEV1 levels and subsequent all-cause mortality was assessed. RESULTS: The highest age-adjusted mortality rate (91.9 deaths per 1000 person-years) was noted among men characterized by the joint conditions of percent predicted FEV1 < or = 70% and BMI < or = 21 kg/m2. This rate was 4.0 times the mortality rate of a "healthy" reference group characterized by percent predicted FEV1 > 70% and 21 < BMI < 25 kg/m2. This rate ratio is attenuated to 3.2 upon statistical control for measures of current and past smoking behavior. Among the three strata of BMI, statistical interaction is reflected in a heterogeneity of mortality rate differences (49.7, 21.8, -9.6 deaths/person-year, respectively) and rate ratios (2.18, 1.98, .66, respectively) comparing men with percent predicted FEV1 < or = 70% to > 70%. CONCLUSION: Joint loss of pulmonary function and relative weight is predictive of subsequent all-cause mortality in excess of additive or multiplicative effects of each condition separately. Smoking behavior may contribute to this observation.

Aged↗
PubMed1996

Spectral turbulence versus time-domain analysis of signal-averaged ECG used for the prediction of different arrhythmic events in survivors of acute myocardial infarction.

INTRODUCTION: Spectral turbulence analysis of the signal-averaged ECG (SAECG) combines spectral analysis with statistical evaluation of spectrograms of individual parts of the QRS complex. It has been suggested that it may be superior to conventional time-domain analysis of the SAECG. METHODS AND RESULTS: This study compared the power of conventional time-domain (40 to 250 Hz) and spectral turbulence analyses of SAECG for the prediction of cardiac death, ventricular tachycardia, sudden arrhythmic death, and arrhythmic events (ventricular tachycardia or fibrillation, and/or sudden arrhythmic death) after acute myocardial infarction in 603 patients. The population excluded patients with bundle branch block and other conduction abnormalities. During the first 2 years of follow-up, there were 40 cardiac deaths, 21 cases of ventricular tachycardia, 1 sudden arrhythmic deaths, and 29 arrhythmic events. The positive predictive accuracy of spectral turbulence analysis was significantly higher than time-domain analysis for cardiac death at most levels of sensitivity (e.g., 26% vs 20% at 40% sensitivity, P < 0.05). The positive predictive accuracies of the two techniques were not statistically different for the prediction of ventricular tachycardia. For the prediction of sudden arrhythmic death and arrhythmic events, the positive predictive accuracy of spectral turbulence was better than that of time-domain analysis only at the higher levels of sensitivity (9% vs 2%, P < 0.001 for sudden arrhythmic death at 60% sensitivity, and 14% vs 11%, P < 0.05 for arrhythmic events at 60% sensitivity). CONCLUSIONS: Spectral turbulence analysis is essentially equivalent to time-domain analysis for the prediction of arrhythmic events after myocardial infarction. However, it performed significantly better than time-domain analysis for the prediction of cardiac death.

Acute Disease↗
PubMed2006

Steiner cephalometric analysis: predicted and actual treatment outcome compared.

OBJECTIVE: To examine the accuracy and precision of the Steiner prediction cephalometric analysis. SETTING AND SUBJECTS: The sample consisted of 275 randomly selected patients, treated between 1970 and 1995 at a university department. METHODS: Lateral cephalograms before (T1) and after orthodontic treatment (T2) were analyzed using the Steiner analysis. A prediction of the final outcome at T2 for the variables ANB degrees, U1 to NA mm, L1 to NB mm, and Pg to NB mm was performed at T1. The difference between the actual outcome at T2 and the Steiner predicted value (SPV), which was done at T1, was calculated. Accuracy (mean difference between T2 and SPV) and precision (standard deviation of the mean prediction discrepancies) of the prediction were studied. Paired t-test was used to detect under- or overestimation of the predicted values. RESULTS: The mean decrease in angle ANB was 1.4 +/- 2.7 degrees and for U1 to NA 2.0 +/- 2.6 mm, while L1 to NB increased 0.8 +/- 2.0 mm and Pg to NB 0.7 +/- 1.1 mm. The predicted values for the changes in ANB angle, the distance of upper incisor U1 to NA as well as the distance Pg to NB were significantly overestimated when compared with the actual outcome, while the change in the distance of lower incisor L1 to NB was underestimated. CONCLUSION: The prediction of cephalometric treatment outcome as used in the Steiner analysis is not accurate enough to base orthodontic treatment decisions upon.

Adolescent↗
PubMed1998

Prediction of serum vancomycin concentrations using one-, two- and three-compartment models with implemented population pharmacokinetic parameters and with the Bayesian method.

Although previous studies have shown that vancomycin has a complicated pharmacokinetic profile requiring description using a two- or, better, three-compartment model, until recently predictions of serum vancomycin concentrations have been mainly based on one- or two-compartment models using computer software packages. In this study, we have predicted serum vancomycin concentrations in 59 patients using one-, two- and three-compartment models with implemented population pharmacokinetic parameters in the Abbott PKS program and by use of the Bayesian method. The percentage errors of predictions made using the one-compartment model were smaller when either the Bayesian method or implemented population pharmacokinetic parameters were used (medians of -8.61% and -9.49%, respectively). Predictions using the one-compartment model with the Bayesian method were less biased (median of -1.52 microgmL(-1). The best predictions were those made using the three-compartment model with the Bayesian method-they were most accurate (median of 3.40 microgmL(-1) and highly precise (median of 11.53 microg(2)mL(-1)). The results suggest that predictions made using the one-compartment model with implemented population pharmacokinetic parameters are preferable if no samples are available, otherwise predictions made using the three-compartment model with the Bayesian method are preferable. The results also supported our previous argument that the greater the number of compartments involved in individualization, the better the predictions obtained using the Bayesian method.

Adult↗
PubMed2006

Experimental evaluation of radiosity for room sound-field prediction.

An acoustical radiosity model was evaluated for how it performs in predicting real room sound fields. This was done by comparing radiosity predictions with experimental results for three existing rooms--a squash court, a classroom, and an office. Radiosity predictions were also compared with those by ray tracing--a "reference" prediction model--for both specular and diffuse surface reflection. Comparisons were made for detailed and discretized echograms, sound-decay curves, sound-propagation curves, and the variations with frequency of four room-acoustical parameters--EDT, RT, D50, and C80. In general, radiosity and diffuse ray tracing gave very similar predictions. Predictions by specular ray tracing were often very different. Radiosity agreed well with experiment in some cases, less well in others. Definitive conclusions regarding the accuracy with which the rooms were modeled, or the accuracy of the radiosity approach, were difficult to draw. The results suggest that radiosity predicts room sound fields with some accuracy, at least as well as diffuse ray tracing and, in general, better than specular ray tracing. The predictions of detailed echograms are less accurate, those of derived room-acoustical parameters more accurate. The results underline the need to develop experimental methods for accurately characterizing the absorptive and reflective characteristics of room surfaces, possible including phase.

Acoustics↗
PubMed2004

Validation of predictive rules and indices of severity for community acquired pneumonia.

BACKGROUND: A study was undertaken to validate the modified American Thoracic Society (ATS) rule and two British Thoracic Society (BTS) rules for the prediction of ICU admission and mortality of community acquired pneumonia and to provide a validation of these predictions on the basis of the pneumonia severity index (PSI). METHOD: Six hundred and ninety six consecutive patients (457 men (66%), mean (SD) age 67.8 (17.1) years, range 18-101) admitted to a tertiary care hospital were studied prospectively. Of these, 116 (16.7%) were admitted to the ICU. RESULTS: The modified ATS rule achieved a sensitivity of 69% (95% CI 50.7 to 77.2), specificity of 97% (95% CI 96.4 to 98.9), positive predictive value of 87% (95% CI 78.3 to 93.1), and negative predictive value of 94% (95% CI 91.8 to 95.8) in predicting admission to the ICU. The corresponding predictive indices for mortality were 94% (95% CI 82.5 to 98.7), 93% (95% CI 90.6 to 94.7), 49% (95% CI 38.2 to 59.7), and 99.5% (95% CI 98.5 to 99.9), respectively. These figures compared favourably with both the BTS rules. The BTS-CURB criteria achieved predictions of pneumonia severity and mortality comparable to the PSI. CONCLUSIONS: This study confirms the power of the modified ATS rule to predict severe pneumonia in individual patients. It may be incorporated into current guidelines for the assessment of pneumonia severity. The CURB criteria may be used as an alternative tool to PSI for the detection of low risk patients.

Adult↗
PubMed2006

Prediction of iatrogenic pseudoaneurysm after percutaneous endovascular procedures.

PURPOSE: To prospectively evaluate the accuracy of using physical examination to identify puncture-related groin pseudoaneurysms, as assessed by using duplex ultrasonography (US), after percutaneous transluminal procedures and to prospectively evaluate the association between preinterventional platelet count, antiplatelet medication, and the occurrence of pseudoaneurysms. MATERIALS AND METHODS: This study was approved by the local ethics committee, and informed consent was obtained from all patients. The study prospectively included 273 consecutive patients (161 men, 112 women; age range, 34-90 years) who were referred for duplex US evaluation of the inguinal arterial puncture site 1 day after endovascular procedures. Prior to duplex US, all patients underwent physical examination of the groin. In addition, clinical characteristics and preinterventional laboratory parameters were assessed. Statistical significance was determined by using chi2 tests, the Fischer exact test, and unpaired t tests. RESULTS: Twenty-three pseudoaneurysms were found in 273 patients by using duplex US. Pulsatile groin masses that were detected at physical examination were used to correctly identify all pseudoaneurysms (positive predictive value, 100%; negative predictive value, 100%). Painful pulse palpation had a slightly lower predictive power (positive predictive value, 92% [95% confidence interval: 81%, 100%]; negative predictive value, 100% [95% confidence interval: 100%, 100%]). Other clinical parameters, such as the presence of superficial hematomas, systolic bruits, or nonpulsatile groin masses, had no adequate predictive properties. Interobserver agreement was excellent between observers (97% agreement [95% confidence interval: 92%, 100%]). All patients with pseudoaneurysms had a preprocedural platelet count of less than 200 x 10(9)/L. No subacute complications were observed at the access site in patients with a platelet count of more than 200 x 10(9)/L. CONCLUSION: Physical examination revealed sufficient predictive capability in facilitating the identification of iatrogenic pseudoaneurysms after percutaneous vascular procedures. A platelet count of less than 200 x 10(9)/L was associated with high predictive capability, thereby warranting further assessment in a larger series of patient.

Adult↗
PubMed1999

Artificial neural networks applied to survival prediction in breast cancer.

In this study, we evaluated the accuracy of a neural network in predicting 5-, 10- and 15-year breast-cancer-specific survival. A series of 951 breast cancer patients was divided into a training set of 651 and a validation set of 300 patients. Eight variables were entered as input to the network: tumor size, axillary nodal status, histological type, mitotic count, nuclear pleomorphism, tubule formation, tumor necrosis and age. The area under the ROC curve (AUC) was used as a measure of accuracy of the prediction models in generating survival estimates for the patients in the independent validation set. The AUC values of the neural network models for 5-, 10- and 15-year breast-cancer-specific survival were 0.909, 0.886 and 0.883, respectively. The corresponding AUC values for logistic regression were 0.897, 0.862 and 0.858. Axillary lymph node status (N0 vs. N+) predicted 5-year survival with a specificity of 71% and a sensitivity of 77%. The sensitivity of the neural network model was 91% at this specificity level. The rate of false predictions at 5 years was 82/300 for nodal status and 40/300 for the neural network. When nodal status was excluded from the neural network model, the rate of false predictions increased only to 49/300 (AUC 0. 877). An artificial neural network is very accurate in the 5-, 10- and 15-year breast-cancer-specific survival prediction. The consistently high accuracy over time and the good predictive performance of a network trained without information on nodal status demonstrate that neural networks can be important tools for cancer survival prediction.

Adult↗
PubMed1998

Can total urinary protein measurements predict microalbuminuria?

We re-addressed the question of whether routine total urinary protein determinations can be used to predict the presence of microalbuminuria by studying 61 patients who attended a diabetic clinic and tested negative or had one positive protein by dipstick. Total urinary protein was measured by the Biorad dye-binding method in undialyzed urine (UND), in dialyzed urine (DIAL), and in dialyzed urine in which albumin and globulins were separated, measured separately with albumin and globulin standards and the results added together to obtain total urinary protein (A + G). The results were compared with albumin measurements obtained by radioimmunoassay (RIA). Compared to DIAL, urinary protein measurements were 43% higher with A + G and 22% higher with UND. Microalbuminuria correlated moderately with UND (r =0.81) and better with the other methods (r=0.87 for DIAL, r=0.91 for A + G). None of the methods predicted microalbuminuria reliably. Taking a protein-to-creatinine ratio of 0.15 and an albumin-to-creatinine ratio of 0.03 as upper limits of normal, we found that UND had a 72% positive predictive value (28% false positives) and 85 % negative predictive value (15% false negatives). DIAL had 90% positive predictive value (10% false positives) and 78% negative predictive value (22% false negatives). A + G had 65% positive predictive value (35% false positives) but 91% negative predictive value (9% false negatives). A + G, which uses the correct standards, would be the most suitable method for screening, having the least number of false negatives, but has more false positives because it is more sensitive. In practice, most routine chemical laboratories find it expedient to use only UND, but physicians interpreting the results of this method should be aware of its limitations.

Albuminuria↗
PubMed2006

Prediction of intra-twin birth weight discordance by binary logistic regression analysis.

AIMS: Identification of women at high risk of intra-twin birth weight discordance is helpful in obstetric care of these pregnancies. The aim of this study is to establish an intra-twin birth weight discordance prediction model. METHODS: We created an intra-twin birth weight discordance prediction model by logistic regression, based on the 1995-1997 register twin birth data of the USA. The twin sets were randomly divided into two groups: group 1 to establish the prediction model and group 2 to validate the prediction model. Intra-twin birth weight discordance was defined as birth weight discordance > 25%. The prediction model was validated by receiver operating characteristic curve. RESULTS: A birth weight discordance prediction model including maternal age (beta = 0.069), parity (beta = 0.250), fetal gender concordance (beta = 0.041), maternal hypertension (beta = 0.368), eclampsia (beta = 0.316), other medical complication (beta = 0.165), and smoking (beta = 0.164) was established, yielded a 0.558 area under the receiver operating characteristic curve. The sensitivity, specificity, and positive predictive values were 38.1, 69.7, and 10.8%, respectively, at the cut-off value of 0.09 in group 2. CONCLUSION: A birth weight discordance prediction model that includes seven variables available during pregnancy has been established with acceptable diagnostic performance.

Area Under Curve↗
PubMed1995

Significance and incidence of concordance of drug efficacy predictions by Holter monitoring and electrophysiological study in the ESVEM Trial. Electrophysiologic Study Versus Electrocardiographic Monitoring.

BACKGROUND: Selection of antiarrhythmic therapy may be based on either suppression of spontaneous ventricular arrhythmias assessed by Holter monitoring or by suppression of inducible ventricular arrhythmias during electrophysiological study. This study examines the frequency and significance of concordance of these two approaches in the Electrophysiologic Study Versus Electrocardiographic Monitoring (ESVEM) trial. METHODS AND RESULTS: Twenty-four-hour Holter monitoring was performed in patients randomized to the electrophysiology limb of the ESVEM study at the time of the first drug trial and at the time of an effective drug trial. Holter monitors were available in 65% (146/226) of patients at the time of the first drug trial and in 93% (100/108) of patients at the time of an electrophysiology study predicting drug efficacy. There were no clinical differences between patients who had and those who did not have a Holter monitor. At the time of the first drug trial, concordance of Holter and electrophysiological predictions of drug efficacy was observed in 46% of patients (both techniques predicted efficacy in 23%; neither predicted efficacy in 23%). Discordant results were observed in 54% (Holter suppression without electrophysiological suppression in 44%; electrophysiological suppression without Holter suppression in 10%). At the time of an electrophysiology study predicting drug efficacy, 68 of the 100 patients without inducible ventricular tachyarrhythmias also had suppression of spontaneous ventricular arrhythmias on the Holter recorded at the time of the electrophysiological study. Neither arrhythmia recurrence nor mortality was significantly different in patients with suppression of both inducible and spontaneous ventricular arrhythmias compared with those with only suppression of inducible arrhythmias. Comparison of patients with suppression of both inducible and spontaneous ventricular arrhythmias with the 188 patients in the Holter limb, in whom efficacy was predicted by Holter monitoring only, revealed no difference in outcome. CONCLUSIONS: In this population, (1) there is frequent discordance in prediction of drug efficacy and inefficacy between electrophysiological study and Holter monitoring; (2) a requirement to fulfill both Holter and electrophysiological efficacy criteria reduces the number of patients with an efficacy prediction; and (3) suppression of both spontaneous ventricular ectopy and inducible ventricular tachyarrhythmias does not identify a group with better outcome.

Aged↗
PubMed2004

Probability of cortical infarction predicted by flumazenil binding and diffusion-weighted imaging signal intensity: a comparative positron emission tomography/magnetic resonance imaging study in early ischemic stroke.

BACKGROUND AND PURPOSE: The differentiation of reversible from irreversible ischemic damage is essential for identifying patients with acute ischemic deficits who may benefit from therapeutic interventions. Diffusion-weighted imaging (DWI) has become the method of choice to detect ischemic lesions. Positron emission tomography (PET) of the central benzodiazepine receptor ligand 11C flumazenil (FMZ) has been shown to be a reliable marker of neuronal integrity. These 2 imaging parameters were compared with respect to the probability to predict cortical infarction in early ischemic stroke. METHODS: In 12 patients with acute stroke, results from DWI (median, 6.5 hours after symptom onset) and FMZ-PET (interval, 85 minutes between DWI and PET) were compared with infarct extension 24 to 48 hours after onset of stroke on T2-weighted magnetic resonance imaging (T2-MRI). Probability curves predictive of eventual infarction were computed using respective DWI, FMZ, and apparent diffusion coefficient (ADC) values for voxels of interest (VOI) later classified as representing infarcted or noninfarcted tissue. RESULTS: Ninety-five percent limits predictive of cortical infarction were determined for relative FMZ binding (< or =3.2), DWI signal intensity (> or =1.18), and ADC values (< or =0.83). Cortical regions with values beyond these 95% limits did not necessarily overlap with nor were fully congruous with final cortical infarct volumes. The respective median volumes for these regions were FMZ median 10.9, range 0 to 99.7 cm3; DWI median 15.2, range 0 to 116.0 cm3; ADC median 12.4, range 0 to 112.7 cm3; and final infarct median 14.9, range 0 to 114.7 cm(3). Overall, 83.5% of the final infarct, on average, was predicted by decreased FMZ binding, 84.7% by increased DWI signal intensity, and 70.9% by a decreased ADC value. The portions of the final infarct not predicted in the early investigation (false-negatives) were 4.8 cm3 (median) for FMZ, 3.7 cm3 for DWI, and 6.0 cm3 for ADC. The false-positive volumes not included in the final infarct were 0 cm3 (median) for FMZ, 5.1 cm3 for DWI, and 3.6 cm3 for ADC. CONCLUSIONS: These results indicate that FMZ-PET and DWI are comparable in the prediction of probability of ischemic cortical infarction, but FMZ-PET carries a lower probability of false-positive prediction. The final infarcts include tissue not identified by these imaging modalities; at the time of the study, these tissue compartments are viable and could benefit from treatment. The discrepancy in predictive probability could be related to the fundamental difference of the measured variables: benzodiazepine receptor activity is a reliable marker of neuronal integrity in the cortex, and movement of water molecules in the extracellular space might be a more variable indicator of tissue damage.

Adult↗
PubMed2004

Profiled support vector machines for antisense oligonucleotide efficacy prediction.

BACKGROUND: This paper presents the use of Support Vector Machines (SVMs) for prediction and analysis of antisense oligonucleotide (AO) efficacy. The collected database comprises 315 AO molecules including 68 features each, inducing a problem well-suited to SVMs. The task of feature selection is crucial given the presence of noisy or redundant features, and the well-known problem of the curse of dimensionality. We propose a two-stage strategy to develop an optimal model: (1) feature selection using correlation analysis, mutual information, and SVM-based recursive feature elimination (SVM-RFE), and (2) AO prediction using standard and profiled SVM formulations. A profiled SVM gives different weights to different parts of the training data to focus the training on the most important regions. RESULTS: In the first stage, the SVM-RFE technique was most efficient and robust in the presence of low number of samples and high input space dimension. This method yielded an optimal subset of 14 representative features, which were all related to energy and sequence motifs. The second stage evaluated the performance of the predictors (overall correlation coefficient between observed and predicted efficacy, r; mean error, ME; and root-mean-square-error, RMSE) using 8-fold and minus-one-RNA cross-validation methods. The profiled SVM produced the best results (r = 0.44, ME = 0.022, and RMSE= 0.278) and predicted high (>75% inhibition of gene expression) and low efficacy (<25%) AOs with a success rate of 83.3% and 82.9%, respectively, which is better than by previous approaches. A web server for AO prediction is available online at http://aosvm.cgb.ki.se/. CONCLUSIONS: The SVM approach is well suited to the AO prediction problem, and yields a prediction accuracy superior to previous methods. The profiled SVM was found to perform better than the standard SVM, suggesting that it could lead to improvements in other prediction problems as well.

Databases, Genetic↗
PubMed2004

Esub8: a novel tool to predict protein subcellular localizations in eukaryotic organisms.

BACKGROUND: Subcellular localization of a new protein sequence is very important and fruitful for understanding its function. As the number of new genomes has dramatically increased over recent years, a reliable and efficient system to predict protein subcellular location is urgently needed. RESULTS: Esub8 was developed to predict protein subcellular localizations for eukaryotic proteins based on amino acid composition. In this research, the proteins are classified into the following eight groups: chloroplast, cytoplasm, extracellular, Golgi apparatus, lysosome, mitochondria, nucleus and peroxisome. We know subcellular localization is a typical classification problem; consequently, a one-against-one (1-v-1) multi-class support vector machine was introduced to construct the classifier. Unlike previous methods, ours considers the order information of protein sequences by a different method. Our method is tested in three subcellular localization predictions for prokaryotic proteins and four subcellular localization predictions for eukaryotic proteins on Reinhardt's dataset. The results are then compared to several other methods. The total prediction accuracies of two tests are both 100% by a self-consistency test, and are 92.9% and 84.14% by the jackknife test, respectively. Esub8 also provides excellent results: the total prediction accuracies are 100% by a self-consistency test and 87% by the jackknife test. CONCLUSIONS: Our method represents a different approach for predicting protein subcellular localization and achieved a satisfactory result; furthermore, we believe Esub8 will be a useful tool for predicting protein subcellular localizations in eukaryotic organisms.

Computational Biology↗