Induction of modular classification rules: using Jmax-pruning

The Prism family of algorithms induces modular classification rules which, in contrast to decision tree induction algorithms, do not necessarily fit together into a decision tree structure. Classifiers induced by Prism algorithms achieve a comparable accuracy compared with decision trees and in some cases even outperform decision trees. Both kinds of algorithms tend to overfit on large and noisy datasets and this has led to the development of pruning methods. Pruning methods use various metrics to truncate decision trees or to eliminate whole rules or single rule terms from a Prism rule set. For decision trees many pre-pruning and postpruning methods exist, however for Prism algorithms only one pre-pruning method has been developed, J-pruning. Recent work with Prism algorithms examined J-pruning in the context of very large datasets and found that the current method does not use its full potential. This paper revisits the J-pruning method for the Prism family of algorithms and develops a new pruning method Jmax-pruning, discusses it in theoretical terms and evaluates it empirically.

Jmax-pruning: a facility for the information theoretic pruning of modular classification rules

The Prism family of algorithms induces modular classification rules in contrast to the Top Down Induction of Decision Trees (TDIDT) approach which induces classification rules in the intermediate form of a tree structure. Both approaches achieve a comparable classification accuracy. However in some cases Prism outperforms TDIDT. For both approaches pre-pruning facilities have been developed in order to prevent the induced classifiers from overfitting on noisy datasets, by cutting rule terms or whole rules or by truncating decision trees according to certain metrics. There have been many pre-pruning mechanisms developed for the TDIDT approach, but for the Prism family the only existing pre-pruning facility is J-pruning. J-pruning not only works on Prism algorithms but also on TDIDT. Although it has been shown that J-pruning produces good results, this work points out that J-pruning does not use its full potential. The original J-pruning facility is examined and the use of a new pre-pruning facility, called Jmax-pruning, is proposed and evaluated empirically. A possible pre-pruning facility for TDIDT based on Jmax-pruning is also discussed.

J-measure based hybrid pruning for complexity reduction in classification rules

Prism is a modular classification rule generation method based on the ‘separate and conquer’ approach that is alternative to the rule induction approach using decision trees also known as ‘divide and conquer’. Prism often achieves a similar level of classification accuracy compared with decision trees, but tends to produce a more compact noise tolerant set of classification rules. As with other classification rule generation methods, a principle problem arising with Prism is that of overfitting due to over-specialised rules. In addition, over-specialised rules increase the associated computational complexity. These problems can be solved by pruning methods. For the Prism method, two pruning algorithms have been introduced recently for reducing overfitting of classification rules - J-pruning and Jmax-pruning. Both algorithms are based on the J-measure, an information theoretic means for quantifying the theoretical information content of a rule. Jmax-pruning attempts to exploit the J-measure to its full potential because J-pruning does not actually achieve this and may even lead to underfitting. A series of experiments have proved that Jmax-pruning may outperform J-pruning in reducing overfitting. However, Jmax-pruning is computationally relatively expensive and may also lead to underfitting. This paper reviews the Prism method and the two existing pruning algorithms above. It also proposes a novel pruning algorithm called Jmid-pruning. The latter is based on the J-measure and it reduces overfitting to a similar level as the other two algorithms but is better in avoiding underfitting and unnecessary computational effort. The authors conduct an experimental study on the performance of the Jmid-pruning algorithm in terms of classification accuracy and computational efficiency. The algorithm is also evaluated comparatively with the J-pruning and Jmax-pruning algorithms.

Influence of light guide tip used in the photo-activation on degree of conversion and hardness of one nanofilled dental composite

The aim of this study was to evaluate the degree of conversion and hardness of a dental composite resin Filtek (TM) Z-350 (3M ESPE, Dental Products St. Paul, MN) photo-activated for 20 s of irradiation time with two different light guide tips, metal and polymer, coupled on blue LED Ultraled LCU (Dabi Atlante, SP, Brazil). With the metal light tip, power density was of 352 and with the polymer was of 456 mW/cm(2), respectively. Five samples (4 mm in diameter and 2mm in thickness-ISO 4049), were made for each Group evaluated. The measurements for DC (%) were made in a Nexus-470 FT-IR, Thermo Nicolet, E.U.A. Spectroscopy (FTIR). Spectra for both uncured and cured samples were analyzed using an accessory of reflectance diffuse. The measurements were recorded in absorbance operating under the following conditions: 32 scans, 4 cm(-1) resolution, 300-4000 cm(-1) wavelength. The percentage of unreacted carbon double bonds (% C=C) was determined from the ratio of absorbance intensities of aliphatic C=C (peak at 1637 cm(-1)) against internal standard before and after curing of the sample: aromatic C-C (peak at 1610 cm(-1)). The Vickers hardness measurements (top and bottom surfaces) were performed in a universal testing machine (Buehler MMT-3 digital microhardness tester Lake Bluff, Illinois USA). A 50 gf load was used and the indenter with a dwell time of 30 s. The data were submitted to the test t Student at significance level of 5%. The mean values of degree of conversion for the polymer and metal light guide tip no were statistically different (p = 0.8389). The hardness mean values were no statistically significant different among the light guide tips (p = 0.6244), however, there was difference between top and bottom surfaces (p < 0.001). The results show that so much the polymer light tip as the metal light tip can be used for the photo-activation, probably for the low quality of the light guide tip metal.

Variability of the conserved V3 loop tip motif in HIV-1 subtype B isolates collected from Brazilian and French patients

The diversity of the V3 loop tip motif sequences of HIV-1 subtype B was analyzed in patients from Botucatu (Brazil) and Montpellier (France). Overall, 37 tetrameric tip motifs were identified, 28 and 17 of them being recognized in Brazilian and French patients, respectively. The GPGR (P) motif was predominant in French but not in Brazilian patients (53.5% vs 31.0%), whereas the GWGR (W) motif was frequent in Brazilian patients (23.0%) and absent in French patients. Three tip motif groups were considered: P, W, and non-P non-W groups. The distribution of HIV-1 isolates into the three groups was significantly different between isolates from Botucatu and from Montpellier (P < 0.001). A higher proportion of CXCR4-using HIV-1 (X4 variants) was observed in the non-P non-W group as compared with the P group (37.5% vs 19.1%), and no X4 variant was identified in the W group (P < 0.001). The higher proportion of X4 variants in the non-P non-W group was essentially observed among the patients from Montpellier, who have been infected with HIV-1 for a longer period of time than those from Botucatu. Among patients from Montpellier, CD4+ cell counts were lower in patients belonging to the non-P non-W group than in those belonging to the P group (24 cells/µL vs 197 cells/µL; P = 0.005). Taken together, the results suggest that variability of the V3 loop tip motif may be related to HIV-1 coreceptor usage and to disease progression. However, as analyzed by a bioinformatic method, the substitution of the V3 loop tip motif of the subtype B consensus sequence with the different tip motifs identified in the present study was not sufficient to induce a change in HIV-1 coreceptor usage.

GIANT FUSIFORM ANEURYSM ARISING FROM FENESTRATED POSTERIOR CEREBRAL ARTERY and BASILAR TIP VARIATION: CASE REPORT

OBJECTIVE: A giant fusiform aneurysm in the posterior cerebral artery (PCA) is rare, as is fenestration of the PCA and basilar apex variation. We describe the angiographic and surgical findings of a giant fusiform aneurysm in the P1-P2 PCA segment associated with PCA bilateral fenestration and superior cerebellar artery double origin.CLINICAL PRESENTATION: A 26-year-old woman presented with a 2-month history of visual blurring. Digital subtraction angiography showed a giant (2.5 cm) fusiform PCA aneurysm in the right P1-P2 segment. The 3-dimensional view showed a caudal fusion pattern from the upper portion of the basilar artery associated with a bilateral long fenestration of the P1 and P2 segments and superior cerebellar artery double origin.INTERVENTION: Surgical trapping of the right P1 -P2 segment, including the posterior communicating artery, was performed by a pretemporal approach. Angiograms performed 3 and 13 months after surgery showed complete aneurysm exclusion, and the PCA was permeated and filled the PCA territory. Clinical follow-up at 14 months showed the patient with no deficits and a return to normal life.CONCLUSION: To our knowledge, this is the first report of a giant fusiform aneurysm of the PCA associated with P1-P2 segment fenestration and other variations of the basilar apex (bilateral superior cerebellar artery duplication and caudal fusion). Comprehension of the embryology and anatomy of the PCA and its related vessels and branches is fundamental to the decision-making process for a PCA aneurysm, especially when parent vessel occlusion is planned.

Comparison between qualitative and semiquantitative catheter-tip cultures: laboratory diagnosis of catheter-related infection in newborns

Este estudo prospectivo avaliou os métodos semiquantitativo e qualitativo de cultura de cateter para o diagnóstico de infecção relacionada a cateter (IRC) em recém-nascidos (RN). Foram incluídas pontas de cateteres provenientes de recém-nascidos internados na Unidade Neonatal do Hospital das Clínicas da Faculdade de Medicina de Botucatu, UNESP. Foram utilizadas as técnicas semiquantitativa e qualitativa de cultura de cateter. Para o diagnóstico de IRC, os microrganismos isolados das culturas de cateteres e de hemoculturas periféricas foram identificados e submetidos ao teste de sensibilidade a antimicrobianos. O padrão ouro correspondeu ao diagnóstico de certeza de IRC, com o isolamento do mesmo microrganismo (espécie e perfil de sensibilidade a antimicrobianos) isolado em hemocultura periférica. Foram estudados 85 cateteres provenientes de 63 RN. A cultura semiquantitativa, embora tenha apresentado menor sensibilidade (90%), apresentou uma maior especificidade (71%) em comparação à sensibilidade de 100% e especificidade de 60% encontradas na cultura qualitativa. Através da identificação dos microrganismos obtidos nas culturas de cateteres, observou-se uma predominância de espécies de Estafilococos coagulase-negativa (ECN). A espécie Staphylococcus epidermidis foi a prevalente (77,5%) nos cateteres com culturas semiquantitativas positivas. Dos 11 episódios de IRC diagnosticados, 8 (72,7%) foram associados a espécies de ECN, dos quais 6 eram da espécie S. epidermidis. Também foram detectados dois casos de IRC por S. aureus e um caso por Candida parapsilosis. O método de cultura semiquantitativo cateter apresentou vantagens para o diagnóstico de IRC em RN quando comparado com o método qualitativo tradicional.

Usefulness of catheter tip culture in the diagnosis of neonatal infections

Objective: To determine the number of colony-forming units (CFU) that best correlates with catheter-related infections (CRI) in newborns.Methods: This was a prospective study of semiquantitative cultures of catheter tips obtained from newborns in the neonatal unit at Faculdade de Medicina de Botucatu, state of São Paulo, Brazil. The microorganisms isolated from catheter and peripheral blood cultures were identified and submitted to a drug susceptibility test. The optimal cutoff point was determined by the receiver operating characteristic (ROC) curve.Results: A total of 85 catheters obtained from 63 newborns were studied. Staphylococcus epidermidis was the predominant species in the catheters (75%). Eight of 11 (72.7%) CRI episodes were associated with coagulase-negative staphylococci, six of which were of the S. epidermidis type. ROC curve analysis indicated that the optimal cutoff point for the diagnosis of CRI was 122 CFU.Conclusions: The cutoff point of 122 CFU correlated best with the diagnosis of CRI in newborns.

Influence of light guide tip used in the photo-activation on degree of conversion and hardness of one nanofilled dental composite

Coordenação de Aperfeiçoamento de Pessoal de Nível Superior (CAPES)

Label-free DNA detection of hepatitis C virus based on modified conducting polypyrrole films at microelectrodes and atomic force microscopy tip-integrated electrodes

We present a new strategy for the label-free electrochemical detection of DNA hybridization for detecting hepatitis C virus based on electrostatic modulation of the ion-exchange kinetics of a polypyrrole film deposited at microelectrodes. Synthetic single-stranded 18-mer HCV genotype-1-specific probe DNA has been immobilized at a 2,5-bis(2-thienyl)-N-(3-phosphoryl-n-alkyl)pyrrole film established by electropolymerization at the previously formed polypyrrole layer. HCV DNA sequences (244-mer) resulting from the reverse transcriptase-linked polymerase chain reaction amplification of the original viral RNA were monitored by affecting the ion-exchange properties of the polypyrrole film. The performance of this miniaturized DNA sensor system was studied in respect to selectivity, sensitivity, and reproducibility. The limit of detection was determined at 1.82 x 10(-21) mol L-1. Control experiments were performed with cDNA from HCV genotypes 2a/c, 2b, and 3 and did not show any unspecific binding. Additionally, the influence of the spacer length of 2,5-bis(2-thienyl)-N-(3-phosphoryl-n-alkyl)pyrrole on the behavior of the DNA sensor was investigated. This biosensing scheme was finally extended to the electrochemical detection of DNA at submicrometer-sized DNA biosensors integrated into bifunctional atomic force scanning electrochemical microscopy probes. The 18-mer DNA target was again monitored by following the ion-exchange properties of the polypyrrole film. Control experiments were performed with 12-base pair mismatched sequences.

Ineffectiveness of pruning to control citrus huanglongbing caused by Candidatus Liberibacter americanus

The huanglongbing (HLB) disease of citrus trees, caused by Candidatus Liberibacter asiaticus and Ca. Liberibacter americanus, was first reported in Brazil in March, 2004. The presence of the disease has caused serious concerns among growers. Pruning experiments were conducted to determine if removal of symptomatic branches or the entire canopy (decapitation) would eliminate infected tissues and save HLB-affected trees. Pruning was done in five blocks on a total of 592 3- to 16 year-old 'Valencia', 'Hamlin' or 'Pera' sweet orange trees showing no symptoms or with two levels of symptom severity. Ten decapitated trees per block were caged and all trees were treated with insecticides to control the psyllid vector, Diaphorina citri. Mottled leaves reappeared on most symptomatic (69.2%) as well on some asymptomatic (7.6%) pruned trees, regardless of age, variety, and pruning procedure. Presence of the pathogen (Ca. Liberibacter americanus) in all symptomatic trees was confirmed by PCR. In general, the greater the symptom severity before pruning the lower the percentage of trees that remained asymptomatic after pruning.

Effect of handpiece tip design on the cutting efficiency of an air abrasion system

Purpose : the aim of this study was to evaluate the effect of nozzle angle and tip diameter on the cutting efficiency of an air abrasion system. Materials and Methods: Thirty-six extracted human third molars were air-abraded with the PrepStar microabrasion machine using a handpiece with either 80degrees or 45degrees nozzle angles with 0.38 or 0.48 nun tip orifice diameters. The following parameters were held constant: abrasive particle size (27 mum), air pressure (80 psi), distance (2 mm.) and duration (15 seconds). The cutting efficiency was compared using enamel, dentin and cementum substrates. Width and depth of the cutting patterns were analyzed and measured using scanning electron micrographs. Results: Statistical analysis using three-way ANOVA and Duncan's Multiple Range test revealed that the width of the cuts was significantly greater when the cavities were prepared using the 45degrees nozzle angle. Significantly deeper cavities were produced with the 80degrees nozzle angle. The tip orifice of the nozzle influenced the cutting efficiency in softer substrates, dentin and cementum. Precise removal of hard tissue is best accomplished using the 80degrees angle nozzle tips for all types of tooth surfaces, enamel, dentin and cementum.

Evaluation of cutting patterns produced with air-abrasion systems using different tip designs.

This study assessed cavity preparations produced with different air abrasion tip parameters. Twelve test groups of extracted teeth were prepared to evaluate the parameters of 80 degrees or 45 degrees nozzle angles and 0.38 or 0.48 mm inner tip diameters. All other factors were held constant. A device was made to hold the specimen and air abrasion handpiece that standardized the distance and position relative to the tooth and time of application. The cavities were evaluated by assessing the rounding of the cavosurface margins and cavity floor. Measurements of cavosurface angles and the angle of concavity were made at the deepest portion of the abraded surface using scanning electron micrographs. The cavosurface angles were compared using paired t-test, and the effects of the tip design parameters were analyzed by ANOVA and Duncan's Multiple Range test. From the cavity patterns found in this study, the authors suggest that 80 degrees angle tips are more appropriate than 45 degrees angle tips for making narrow, deep cuts for preventive resin restorations. Conversely, when shallow preparations are needed, as in the case of Class V cavity preparations, cutting patterns of 45 degrees angle tips are more suitable.