Multivariate Stepwise Discriminant Analysis Research Articles

Characterization of Almond Cultivars by the Use of Thermal Analysis Techniques. Application to Cultivar Authenticity

AbstractAlmonds are subjected to thermal processes in the production of processed food and this can affect their thermal stability and lead to oxidation processes. In this work, almond samples from three different cultivars (Spanish Guara and Marcona, and American Butte) were characterized by using Differential Scanning Calorimetry (DSC) and Thermogravimetric Analysis (TGA) at different heating rates. Crystallization and melting parameters were determined by DSC; whereas thermal stability was studied by TGA, showing no apparent degradation for all samples up to around 290 °C. Butte samples showed the lowest DSC values and TGA initial degradation temperature. These results were linked with differences in fatty acid profiles between Butte and Spanish almond cultivars, Butte presenting higher linoleic acid content. Successful discrimination was obtained for samples analyzed at 2 and 10 °C min−1 heating rates for DSC and TGA, respectively, by applying multivariate stepwise linear discriminant analysis (LDA). The results obtained proved the suitability of thermal analysis techniques combined with LDA for an easy and fast discrimination among different almond cultivars to control eventual adulteration in food processing.

Monitoring the oxidation of almond oils by HS-SPME–GC–MS and ATR-FTIR: Application of volatile compounds determination to cultivar authenticity

In this study, the oxidative process of Spanish and American almond oils was monitored by using headspace solid-phase microextraction/gas chromatography–mass spectrometry (HS-SPME/GC–MS) and attenuated total reflectance Fourier transform infrared spectroscopy (ATR-FTIR). Oxidative stability of samples was evaluated after thermal treatment at 100°C for 1, 3, 5, 7, 10, 15 and 20days, in order to accelerate the lipid oxidation process. Variations observed in bands of infrared spectra were used to monitor the progress of the oxidation process of almond oils. HS-SPME–GC–MS was used for isolation and determination of volatile compounds in oxidised almond oils. The following aldehydes were identified in all samples: hexanal, (E)-2-heptenal, (E)-2-octenal, nonanal, (E)-2-nonenal, (E,E)-2,4-nonadienal and (E,E)-2,4-decadienal. Determination of volatile compounds resulting from lipid oxidation from three different almond cultivars (Spanish Guara and Marcona and Butte from USA) was performed to obtain a set of parameters for discrimination between Spanish and American cultivars. Successful discrimination was obtained in samples kept under thermal treatment for 7days, by using HS-SPME/GC–MS combined with multivariate stepwise linear discriminant analysis (LDA).

Drastic and synchronous changes in megabenthic community structure concurrent with environmental variations in a eutrophic coastal bay

This study demonstrated long-term changes in the stock size and species composition of the megabenthic community, including both vertebrates and invertebrates, in Tokyo Bay, Japan, a semi-enclosed coastal area in which the ecosystem has been affected by both natural and anthropogenic factors. We conducted fisheries-independent trawl surveys over a total of 26 yr (1977–1995 and 2003–2009) to collect megabenthos from the entire bay. We also obtained environmental data for the bay from published records to investigate relationships between biota and environment. Abundance- and biomass-based densities of the megabenthos increased until the late 1980s, followed by an abrupt decline, which we attribute to changes in the densities of small to mid-sized dominant species. In the 2000s, although the abundance-based density remained low, biomass-based density increased substantially due to an increase in the numbers of large fishes. Cluster analysis divided the survey period into six groups showing characteristic species compositions and densities. Stepwise multivariate discriminant analysis (SMDA) revealed that changes in water temperature, dissolved oxygen concentration, precipitation, and copepod density were accompanied by simultaneous changes in the megabenthic community. The SMDA cross-validation test indicated that the percentage of incorrect classifications of the model ranged from 14.3% to 66.7%, implying that factors not incorporated in the model, such as fisheries pressure, might affect the changes in the megabenthic community structure. Our results suggest that major shifts have occurred in the megabenthic community structure in Tokyo Bay, and that both natural and anthropogenic factors can trigger these shifts in the system.

Forecasting business failure in China using case‐based reasoning with hybrid case respresentation

AbstractCase‐based reasoning (CBR) is a very effective and easily understandable method for solving real‐world problems. Business failure prediction (BFP) is a forecasting tool that helps people make more precise decisions. CBR‐based BFP is a hot topic in today's global financial crisis. Case representation is critical when forecasting business failure with CBR. This research describes a pioneer investigation on hybrid case representation by employing principal component analysis (PCA), a feature extraction method, along with stepwise multivariate discriminant analysis (MDA), a feature selection approach. In this process, sample cases are represented with all available financial ratios, i.e., features. Next, the stepwise MDA is used to select optimal features to produce a reduced‐case representation. Finally, PCA is employed to extract the final information representing the sample cases. All data signified by hybrid case representation are recorded in a case library, and the k‐nearest‐neighbor algorithm is used to make the forecasting. Thus we constructed a hybrid CBR (HCBR) by integrating hybrid case representation into the forecasting tool. We empirically tested the performance of HCBR with data collected for short‐term BFP of Chinese listed companies. Empirical results indicated that HCBR can produce more promising prediction performance than MDA, logistic regression, classical CBR, and support vector machine. Copyright © 2009 John Wiley & Sons, Ltd.

Screening for obstructive sleep apnea/hypopnea syndrome: Subjective and objective factors

Objectives To determine the sensitivity and specificity of the Berlin Questionnaire and the Obstructive Sleep Apnea/Hypopnea Syndrome (OSAHS) score for individually predicting a diagnosis of OSAHS, and to propose a method for OSAHS screening incorporating objective and subjective factors. Study Design Cross-sectional survey. Setting Tertiary care center. Methods Charts were reviewed from 223 consecutive patients for whom complete data regarding the Berlin questionnaire, OSAHS score (Friedman tongue position + tonsil size + body mass index grade), Epworth Sleepiness Scale, and visual analog scale for snoring were obtained prior to polysomnography (PSG). Sensitivity and specificity were determined for the Berlin questionnaire and OSAHS score for predicting an apnea hypopnea index (AHI) ≥ 5. Patient data were subjected to multivariate stepwise discriminant analysis and used to construct a screening system based on the Fisher's linear classification equation. Results were cross-validated by PSG findings. Results In predicting an AHI ≥ 5, the sensitivity and specificity, respectively, were 0.615 and 0.226 for the Berlin questionnaire, 0.863 and 0.468 for OSAHS score, and 0.82 and 0.834 for our predictive equation. When applied case-wise to the study population, this equation correctly predicted 82.5 percent of diagnoses. Accuracy was highest for severe OSAHS (87.4%) and lowest for mild disease (77.0%). Sensitivity was lowest for mild OSAHS (0.50). Conclusions Neither the Berlin questionnaire nor the OSAHS score alone was both highly sensitive and specific for diagnosing OSAHS. By incorporating subjective and objective metrics into a single predictive equation, sensitivity and specificity were maximized, and 82.5 percent of diagnoses were accurately predicted.

The diagnostic value of seminal adenosine triphosphate (ATP) in an in vitro fertilization (IVF) program: Der diagnostische Wert von ATP in einem IVF-Programm

The level of adenosine triphosphate (ATP) was quantitated in semen samples used for in vitro fertilization of human oocytes. Seminal ATP level correlated with the concentration and percentage motility of spermatozoa but not with the in vitro fertilization rate of human oocytes. Seminal ATP measurement appears to have little diagnostic value in predicting the fertilizing capacity of spermatozoa as evaluated by the multivariate stepwise discriminant analysis.

Puberty timing and fluid intelligence: A study of correlations between testosterone and intelligence in 8- to 12-year-old Chinese boys

Sex hormone such as testosterone was recently recognized as an important contributor of spatial cognition and intelligence during development, but the relationship between puberty timing and intelligence especially in children is largely unknown. Here in this study, we investigated the potential relationship between the level of sex hormones in saliva and fluid intelligence in 8- to 12-year-old Chinese boys. Fluid intelligence was measured by the Cattell's Culture Fair Intelligence Test. 1600 children aged 8-12 years were included in the Cattell's Culture Fair Intelligence Test and saliva samples were collected thereafter from 166 boys with normal intelligence distribution, composed of 49, 54 and 63 boys in 8-, 10- and 12-year-old group respectively. The level of salivary testosterone and estradiol was measured with enzyme-immunoassay technique. Data of BMI and age were collected. The relationship between the level of salivary sex hormones and fluid intelligence was analysed by correlation test. There was no significant correlation between salivary testosterone level and fluid intelligence in 8-year-old boys, whereas there was a significant positive correlation in 10-year-old boys and a significant negative correlation in 12-year-old boys between those two variable. To verify the correlation, we performed stepwise multivariate linear regression and discriminant analysis, with both the age and BMI of the boys and their parents, and salivary estradiol level considered. The results showed that the level of testosterone and intelligence was correlated, and the correlation was much stronger when the level of salivary testosterone was higher than 14 pg/ml. In summary, the study suggests that the relationship of testosterone and intelligence varies from late childhood to early adolescence, and the puberty timing is closely related with fluid intelligence.

Gene Expression Profiling of Myeloma Cells To Predict Response to Primary Therapy with Thalidomide-Dexamethasone for Newly Diagnosed Multilple Myeloma.

Background: Thalidomide which represents an effective treatment strategy for relapsed/refractory multiple myeloma, actually represents a standard of care also for newly diagnosed multiple myeloma patients.Methods: In the present study, we adopted a gene expression profiling (GEP) strategy in an attempt to predict response (> 50% reduction in serum M protein) to primary therapy with thalidomide-dexamethasone for newly diagnosed multiple myeloma. Plasma cells (CD138+) were purified from bone marrow aspirates from 17 patients at diagnosis, before initiation of treatment with thalidomide-dexamethasone. GEP was performed using the Affymetrix U133 Plus_2 microarray platform. The Affymetrix output (CEL files) was imported into Genespring 7.3 (Agilent technologies) microarray analysis software, where data files were normalized across chips using GCRMA and to the 50th percentile, followed by per gene normalization to median. Criteria of response were those established by Bladè et al.Results: After sufficient follow-up, responders (n=9) and nonresponders (n=8) were identified, and gene expression differences in baselines samples were examined. Of the 11000 genes surveyed, Wilcoxon rank sum test identified 149 genes that distinguished response from non response. A multivariate step-wise discriminant analysis (MSDA) revealed that 14 of the 149 genes could be used in a response predictor model (see table). Of interest, the gene list encompasses WXSC1, known to be involved in the chromosomal translocation t(4;14) (p16.3;q32.3) in multiple myeloma.Conclusion: These results could be the first step to adopt microfluidic cards, in an attempt to select at diagnosis patients who will respond favourably to a particular treatment strategy.List of 14 genes able to predict response to primary therapy with thalidomide-dexamethasone for newly diagnosed multiple myeloma.Gene IDGene NameChromosomal location212771_atC10orf3810p13229874_x_atLOC4007411p36.13219690_atU2AF1L419q13.12202207_atARL72q37.1243819_atGNG214q21203753_atTCF418q21.1235400_atFCRLM11q23.3211474_s_atSERPINB66p25226785_atATP11CXq27.1215440_s_atBEXL1Xq22.1–q22.3209054_s_atWXSC14p16.3227168_atFLJ2596722p12.1213355_atST3GAL63q12.1223218_s_atNFKBIZ3p12–q12

A Gene Expression-Based Risk Stratification Model Developed in Newly Diagnosed Multiple Myeloma Treated with High Dose Therapy Is Predictive of Outcome in Relapsed Disease Treated with Single Agent Bortezomib.

Using data derived from the U133Plus2.0 microarray (U2), we recently constructed a 17-gene model predictive of high-risk multiple myeloma (MM). In the model, 13% of newly diagnosed cases were considered to have high-risk MM with 24 month overall survival estimates of 50% and 90% in the high- and low-risk groups, respectively (p<0.0001). Although validated in a separate cohort of newly diagnosed cases also receiving high dose therapy, it is not known whether this model has broader utility. To address this issue we applied the model to a dataset from a pharmacogenomic effort as part of a multi-center phase III trial testing the efficacy of the proteasome inhibitor bortezomib compared to dexamethasone. For purposes of evaluating the robustness of our model, this dataset is particularly useful in that it differs from the UAMS dataset in several important areas, such as relapsed vs. newly diagnosed disease, single agent bortezomib or dexamethasone vs. multi-agent induction therapy and high-dose melphalan-based tandem transplants, multiple centers vs. single center, delayed processing of bone marrow aspirates vs. immediate processing, negative vs. positive selection of plasma cells, and microarray platform U133A/B (UA) vs. U2 in the Millennium and UAMS datasets, respectively. A total of 144 of the first 351 UAMS cases applied to the U2 had also been applied to UA and were analyzed together with the 188 Millennium samples. Of the 17 genes on U2 we were able to find exact matches for 16 genes on UA. We applied the UA signal intensities of the 16 genes and the exact multivariate stepwise discriminant analysis (MSDA) model used to derive the 17-gene model using U2 data to the 144 UAMS cases. Risk scores for the 144 UAMS cases using the 17-gene U2- and 16-gene UA-derived models were calculated and a strong correlation observed (r=0.89; P<0.001). Using a score of greater than 1.6 as the high-risk cut-point with both models, a confusion matrix revealed a 96.5% agreement between high- and low-risk annotations among the 144 cases. Kaplan-Meier survival analysis of high- and low-risk groups defined by the 17- and 16-gene models revealed similar survival differences between high- and low-risk MM (U2: P=0.0046, HR=2.49; UA: P=0.0026; HR=2.58). These data suggests that the risk model is highly stable across these platforms. We then applied the 16-gene model to 188 cases of relapsed disease treated with bortezomib. Using the same discriminant score cut-off of 1.6 on the Millennium data, the 16-gene model defined 17.6% of cases as high-risk and this subset had significantly shorter overall survival times (P<0.0001; HR=2.93). The estimated 24-month survival was 9% in the high-risk vs. 40% in the low-risk cohort. The model also identified high-risk disease in a dexamethasone treated cohort (P<0.0001; HR=3.01). These data suggest that the 17-gene high-risk model is robust and not confounded by variables distinguishing these two datasets. These data also reveal a common molecular signature of high-risk MM in patients with newly diagnosed and relapsed disease pointing to a potential common molecular mechanism of drug resistance in MM that is revealed in this signature. We are currently developing and validating the 17-gene model using technology that quantifies tumor RNA directly in purified cells.

Time Course of Early Metabolic Changes following Diffuse Traumatic Brain Injury in Rats as Detected by1H NMR Spectroscopy

Experimental models of traumatic brain injury (TBI) provide a useful tool for understanding the cerebral metabolic changes induced by this pathological condition. Here, we report on the time course of changes in cerebral metabolites after TBI and its correlation with early brain morphological changes using a combination of high-resolution proton magnetic resonance spectroscopy ((1)H MRS) and magnetic resonance imaging (MRI). Adult male Sprague-Dawley rats were subjected to closed head impact and examined by MRI at 1, 9, 24, 48, and and 72 h after the injury. Extracts from funnel frozen rat brains were then obtained and analyzed quantitatively by high-resolution (1)H MRS. Finally, statistical multivariate analysis was carried out to identify the combination of cerebral metabolites that best described the time evolution of diffuse TBI. The temporal changes observed in the concentration of cerebral metabolites followed three different patterns. The first pattern included taurine, threonine, and glycine, with concentrations peaking 24 h after the injury. The second pattern included glutamate, GABA, and alanine, with concentrations remaining elevated between 24 and 48 h post-injury. The third one involved creatine-phosphocreatine, N-acetylaspartate, and myo-inositol, with concentrations peaking 48 h after the injury. A multivariate stepwise discriminant analysis revealed that the combination of the organic osmolytes taurine and myo-inositol allowed optimal discrimination among the different time groups. Our findings suggest that the profile of some specific brain molecules that play a role as organic osmolytes can be used to follow-up the progression of the early diffuse brain edema response induced by TBI.

An Automated Feature Selection Method for Visual Inspection Systems

Automated visual inspection (AVI) systems these days are considered essential in the assembly of surface-mounted devices (SMDs) in the electronics industry. This industry has faced the problem of rapid introduction and retirement of SMD-based products with the consequent obsolescence of the inspection systems already in the assembly lines. The constant introduction of new products has caused AVI systems to become rapidly obsolete. The general goal of this research centers on developing self-training AVI systems for the inspection of SMD components. The premise is that these systems would be less prone to obsolescence. In this paper, the authors describe the methodology being used for automatically selecting the features to inspect new components. In particular, this paper explores the use of multivariate stepwise discriminant analysis techniques, such as Wilks' Lambda, in order to automate the feature selection process. All of these techniques are applied to a case study of the inspection of SMD components. Note to Practitioners-In this paper, we present a methodology that would allow the automation of the tedious task of exploring and selecting features to develop algorithms for automated visual inspection systems. In particular, the proposed method selects a subset of features among all of the known features. The chosen subset seeks to minimize inspection errors while keeping the algorithmic development time to a minimum. This is particularly useful for adapting pre-existing systems to inspect new components, especially when the characteristics of the new components are similar to those of components already inspected by the inspection system. We applied this methodology to a case of study of the inspection of surface-mounted devices

Dental esthetic self-perception in young adults with and without previous orthodontic treatment.

The aim of this study was to determine which dental esthetic self-perception evaluation tool discriminated better between orthodontically treated or untreated Peruvian young adults. A total of 630 students were randomly chosen from the 2000 admitted in 2002 to a private university in Peru. Students undergoing active orthodontic treatment at the time of examination were excluded. Self-perceived dental esthetic appearance was evaluated through Standardized Continuum of Aesthetic Need, Oral Aesthetic Subjective Index Scale, and Visual Analogue Scale (VAS). A stepwise multivariate discriminant analysis was developed to classify orthodontic treatment according to the three evaluation tools. A total of 199 students (31.6%) reported a history of orthodontic treatment. Differences between orthodontically treated and untreated groups were found only for the mean VAS score (P < .001). Although three different approaches were used to evaluate dental esthetic self-perception, only VAS allowed the discrimination of the self-perceived dental appearance between orthodontically treated and untreated Peruvian university young adults. Similarities in the self-perceived dental appearance of treated and untreated groups reported in previous epidemiological studies could be explained because different evaluation instruments were used. Further studies are required to support current findings.

Reciprocal Changes in 12-Lead Electrocardiography Can Predict Left Main Coronary Artery Lesion in Patients With Acute Myocardial Infarction

Acute left main coronary artery (LMCA) occlusion may result in acute myocardial infarction (AMI) or sudden death. ST elevation in the aVR and V1 leads is reported to be valuable in recognizing LMCA occlusion. Early recognition of electrocardiogram (ECG) changes, such as reciprocal ST depression in other leads, is helpful in averting this disaster. This study aimed to determine the reciprocal ST segment depression of 12-lead ECGs associated with acute LMCA occlusion. From January 2000 to December 2004, 61 patients who underwent emergency percutaneous coronary intervention in 3 hospitals due to AMI associated with LMCA (n = 18) and a left anterior descending coronary artery (LADCA) (n = 43) proximal lesion were selected. Reciprocal ST segment depression occurred in leads aVF, V(2), V(3), V(4), V(5), and V(6) with significantly higher incidence in the LMCA group than in the LADCA group. Stepwise linear multivariate discriminant analysis indicated that ST segment depression in leads aVF, V(2), and V(4) could distinguish the LMCA group from the LADCA group. We concluded that reciprocal ST segment depression in leads V(2), V(4), and aVF of a 12-lead ECG is an important predictor of acute LMCA occlusion.

Strength Around the Hip and Flexibility of Soft Tissues in Individuals With and Without Patellofemoral Pain Syndrome

Case control design. To investigate whether differences exist in lower extremity muscle strength and soft tissue length between patients with patellofemoral pain syndrome (PFPS) and age- and gender-matched control subjects. Based on our clinical experience and emerging data, impairments such as muscular weakness surrounding the hip and limited flexibility of key lower extremity muscles may be important impairments to consider in the conservative management of PFPS. Thirty patients with PFPS and 30 age- and gender-matched controls without PFPS (17 females and 13 males in each group) participated in the study. Data were collected during 1 testing session by an examiner not blinded to group assignment. Demographic, health history, physical activity levels, and pain and function were assessed using patient-completed measures. Physical examination measures included assessment of hip external rotation strength, hip abduction strength, length of the iliotibial band/tensor fascia lata complex, gastrocnemius length, soleus length, and quadriceps and hamstrings muscles length. Patients with PFPS demonstrated significantly less flexibility of the gastrocnemius, soleus, quadriceps, and hamstrings compared to healthy control subjects. No differences existed in flexibility of the iliotibial band/tensor fascia lata complex and strength of the hip external rotators and abductors. Multivariate stepwise discriminant analysis identified gastrocnemius length, hip abduction strength, and soleus length as being able to distinguish between patients with PFPS and healthy individuals without PFPS. This study suggests that further research is warranted regarding the association of these impairments of muscle strength and soft tissue length in patients with PFPS.

Application of multielement stable isotope ratio analysis to the characterization of French, italian, and spanish cheeses.

The stable isotope ratios (delta13C, delta15N, and delta34S of casein and delta13C and delta18O of glycerol) measured by IRMS of French, Italian, and Spanish cheeses are presented and discussed. Variability factors such as animal-feeding regimen, geographical origin, and climatic and seasonal conditions were studied to check the possibilities of cheese characterization offered by each isotopic parameter. Delta13C values of both casein and glycerol appeared to be strongly correlated to the amount of maize in the animal diet. Delta15N and delta34S of casein proved to be mostly influenced by the geoclimatic conditions of the area (aridity, closeness to the sea, altitude). Delta18O of glycerol was more dependent on the geographical origin of the cheeses and on climatic/seasonal parameters. By applying a multivariate stepwise canonical discriminant analysis, good discrimination possibilities for the different European cheeses were obtained, confirmed by the classification analysis, when >90% of the samples were correctly reclassified.

Use of remote sensing and geographical information systems to identify environmental features that influence the distribution of paramphistomosis in sheep from the southern Italian Apennines

A geographic information system (GIS) was constructed using remote sensing (RS) and landscape feature data together with Calicophoron daubneyi positive survey records from 197 georeferenced ovine farms with animals pasturing in a 3971 km 2 area of the southern Italian Apennines. The objective was to study the spatial distribution of this rumen fluke, identify environmental features that influence its distribution, and develop a preliminary risk assessment model. The GIS for the study area was constructed utilizing the following environmental variables: normalized difference vegetation index (NDVI), land cover, elevation, slope, aspect, and total length of rivers. These variables were then calculated for “buffer zones” consisting of the areas included in a circle of 3 km diameter centered on 197 farms. The environmental data obtained from GIS and RS and from data taken by the veterinarians on the field (stocking rate and presence of streams, springs and brooks on pasture) were analyzed by univariate (Spearman and ANOVA) and multivariate (discriminant) statistical analyses using the farm coprological status (positive/negative) as the dependent variable. Sheep on 32 of the 197 (16.2%) farms, were positive for C. daubneyi, with an average intensity of 52 epg. A multivariate stepwise discriminant analysis model was developed that included moors and heathland, sclerophyllous and coniferous forest vegetation, autumn–winter NDVI and presence of streams, springs and brooks on pasture. The variables entered in the model were also correlated with C. daubneyi positive farms in the univariate tests and are consistent with the environmental requirements of C. daubneyi and its snail intermediate host.

Linear discriminant analysis of dermoscopic parameters for the differentiation of early melanomas from Clark naevi.

As a first step to develop a screening system for pigmented skin lesions, we performed digital discriminant analyses between early melanomas and Clark naevi. A total of 59 cases of melanoma, including 23 melanoma in situ and 36 thin invasive melanomas (Breslow thickness < or =0.75 mm), and 188 clinically equivocal, histopathologically diagnosed Clark naevi were used in our study. After calculating 62 mathematical variables related to the colour, texture, asymmetry and circularity based on the dermoscopic findings of the pigmented skin lesions, we performed multivariate stepwise discriminant analysis using these variables to differentiate melanomas from naevi. The sensitivities and specificities of our model were 94.4 and 98.4%, respectively, for discriminating between melanomas (Breslow thickness < or =0.75 mm) and Clark naevi, and 73.9 and 85.6%, respectively, for discriminating between melanoma in situ and Clark naevi. Our algorithm accurately discriminated invasive melanomas from Clark naevi, but not melanomas in situ from Clark naevi.

Behaviour in children with language development disorders.

The objective of the study was to explore the univariate and multivariate differences in behavioural problems among children with disorders in expressive or mixed receptive-expressive language development and children with unimpaired language development. Ninety-four children with language development disorders (LDD) between the ages of 4 and 6 years and 94 children (matched by age and sex) without disorders of language development were compared concerning behavioural problems, as measured by the German version of the Child Behavior Checklist/4-18. Thirty-two children (34%) with LDD showed behavioural problems in the clinical range, whereas only 6 control subjects (6%) had scores in this range. Univariate group comparisons between patients and control subjects showed significant differences in all 8 syndromes and the scale "other problems," with patients having higher scores. Multivariate stepwise discriminant analysis showed a significant discriminant function by the scales "other problems," "social problems," "anxious-depressed," "thought problems," "attention problems," and "delinquent problems." In general, our results agree with several studies that report that children with speech and language disorders are at special risk for developing behavioural problems. Neurodevelopmental immaturity may be one factor underlying both the disorder in language development and the behavioural problems.

Selection of input features across subjects for classifying crewmember workload using artificial neural networks

The issue of crewmember workload is important in complex system operation because operator overload leads to decreased mission effectiveness. Psychophysiological research on mental workload uses measures such as electroencephalogram (EEG), cardiac, eye-blink, and respiration measures to identify mental workload levels. This paper reports a research effort whose primary objective was to determine if one parsimonious set of salient psychophysiological features can be identified to accurately classify mental workload levels across multiple test subjects performing a multiple task battery. To accomplish this objective, a stepwise multivariate discriminant analysis heuristic and artificial neural network feature selection with a signal-to-noise ratio (SNR) are used. In general, EEG power in the 31-40-Hz frequency range and ocular input features appeared highly salient. The second objective was to assess the feasibility of a single model to classify mental workload across different subjects. A classification accuracy of 87% was obtained for seven independent validation subjects using neural network models trained with data from other subjects. This result provides initial evidence for the potential use of generalized classification models in multitask workload assessment.

Predicting revascularization outcome in patients with coronary artery disease and left ventricular dysfunction (data from the SEMINATOR study)

A main goal of revascularization in patients with chronic ischemic cardiomyopathy is to improve global left ventricular (LV) function. This study aimed to verify whether it is possible to predict an increase in LV ejection fraction (EF) after revascularization on the basis of the extent of LV asynergy, myocardial viability, and revascularization completeness. We studied 77 patients with chronic LV ischemic dysfunction using baseline resting and nitrate-enhanced technetium-99m sestamibi single-photon emission computed tomography. Regional wall motion and global LVEF were assessed with echocardiography before and after revascularization, which was complete in 51 patients and incomplete in 26. The number of viable asynergic segments included in revascularized coronary artery territories was the strongest predictor of significant (≥5 EF U) functional improvement in univariate discriminant analysis. According to multivariate stepwise discriminant analysis, this parameter, together with the number of baseline asynergic segments, allowed the detection of patients with significant LVEF improvement with 75% accuracy. With use of a multivariate regression model, including the 2 mentioned variables, the measure of postrevascularization LVEF increase could be accurately quantified (R 2 0.43, p <0.000001). In conclusion, this study suggests that the severity of baseline asynergy, the extent of myocardial viability, and the completeness of revascularization are the main determinants of postrevascularization functional recovery in patients with LV ischemic dysfunction, and that on the basis of these variables it is possible to predict the measure of LVEF increase.