Expectation Propagation with Factorizing Distributions: A Gaussian Approximation and Performance Results for Simple Models

We discuss the expectation propagation (EP) algorithm for approximate Bayesian inference using a factorizing posterior approximation. For neural network models, we use a central limit theorem argument to make EP tractable when the number of parameters is large. For two types of models, we show that EP can achieve optimal generalization performance when data are drawn from a simple distribution.

The impact of simple donor education on donor behavioral deferral and infectious disease rates in Sao Paulo, Brazil

BACKGROUND: Studies have shown that human immunodeficiency virus (HIV) residual risk is higher in Brazilian than in US and European blood donors, probably due to failure to defer at-risk individuals in Brazil. This study assessed the impact of an educational brochure in enhancing blood donors` knowledge about screening test window phase and reducing at-risk individuals from donating. STUDY DESIGN AND METHODS: This trial compared an educational intervention with a blood center`s usual practice. The brochure was distributed in alternating months to all donors. After donating, sampled participants completed two questions about their HIV window period knowledge. The impact on HIV risk deferral, leaving without donation, confidential unit exclusion (CUE) use, and test positivity was also analyzed. RESULTS: From August to November 2007 we evaluated 33,940 donations in the main collection center of Fundacao Pro-Sangue/Hemocentro de Sao Paulo in Sao Paulo, Brazil. A significant (p < 0.001) pamphlet effect was found on correct responses to both questions assessing HIV window phase knowledge (68.1% vs. 52.9%) and transfusion risk (91.1% vs. 87.2%). After adjusting for sex and age, the pamphlet effect was strongest for people with more than 8 years of education. There was no significant pamphlet effect on HIV risk deferral rate, leaving without donation, use of CUE, or infectious disease rates. CONCLUSION: While the educational pamphlet increased window period knowledge, contrary to expectations this information alone was not enough to make donors self-defer or acknowledge their behavioral risk.

Derivation and external validation of a simple prediction model for the diagnosis of type 2 Diabetes Mellitus in the Brazilian urban population

A risk score model was developed based in a population of 1,224 individuals from the general population without known diabetes aging 35 years or more from an urban Brazilian population sample in order to select individuals who should be screened in subsequent testing and improve the efficacy of public health assurance. External validation was performed in a second, independent, population from a different city ascertained through a similar epidemiological protocol. The risk score was developed by multiple logistic regression and model performance and cutoff values were derived from a receiver operating characteristic curve. Model`s capacity of predicting fasting blood glucose levels was tested analyzing data from a 5-year follow-up protocol conducted in the general population. Items independently and significantly associated with diabetes were age, BMI and known hypertension. Sensitivity, specificity and proportion of further testing necessary for the best cutoff value were 75.9, 66.9 and 37.2%, respectively. External validation confirmed the model`s adequacy (AUC equal to 0.72). Finally, model score was also capable of predicting fasting blood glucose progression in non-diabetic individuals in a 5-year follow-up period. In conclusion, this simple diabetes risk score was able to identify individuals with an increased likelihood of having diabetes and it can be used to stratify subpopulations in which performing of subsequent tests is necessary and probably cost-effective.

Six-minute walk test: functional capacity of severely obese before and after bariatric surgery

Background: The 6-minute walk test (6MWT) is a well-known instrument for assessing the functional capacity of a variety of groups, including the obese. It is a simple, low-cost and easily applied method to objectively assess the level of exercise capacity. The aim of the present study was to study the functional capacity of a severely obese population before and after bariatric surgery. Methods: A total of 51 patients were studied. Of the 51 patients, 86.2% were women, and the mean age was 40.9 +/- 9.2 years. All 51 patients were evaluated preoperatively and 49 were evaluated 7-12 months postoperatively. The initial body mass index was 51.1 +/- 9.2 kg/m(2), and the final body mass index was 28.2 +/- 8.1 kg/m(2). All patients underwent Roux-en-Y gastric bypass. The 6MWT was performed in a hospital corridor, with patients attempting to cover as much distance as they could, walking back and forth for as long as possible within 6 minutes at their regular pace. The total distance, Borg Scale of perceived exhaustion, modified Borg dyspnea scale for shortness of breath, and physical complaints at the end of the test were recorded. In addition, the heart rate and respiratory frequency were assessed before and after the test. Results: The tolerance was good, and no injuries occurred at either evaluation. The patients` mean distance for the 6MWT was 381.9 +/- 49.3 m before surgery and 467.8 +/- 40.3 m after surgery (p < .0001). Similar results were observed for the other parameters assessed. Conclusion: The 6MWT provided useful information about the functional status of the obese patients undergoing bariatric surgery. A simple, safe, and powerful method to assess functional capacity of severely obese patients, the 6MWT is an objective test that might replace the conventional treadmill test for these types of patients. (Surg Obes Relat Dis 2009;5:540-543.) (C) 2009 American Society for Metabolic and Bariatric Surgery. All rights reserved.

Frontal Assessment Battery (FAB) is a simple tool for detecting executive deficits in chronic cannabis users

Background: Cannabis is the most used illicit drug in the world, and its use has been associated with prefrontal cortex (PFC) dysfunction, including deficits in executive functions (EF). Considering that EF may influence treatment outcome, it would be interesting to have a brief neuropsychological battery to assess EF in chronic cannabis users (CCU). In the present study, the Frontal Assessment Battery (FAB), a brief, easy to use neuropsychological instrument aimed to evaluate EF, was used to evaluate cognitive functioning of CCU. Methods: We evaluated 107 abstinent CCU with the FAB and compared with 44 controls matched for age, estimated IQ, and years of education. Results: CCU performed poorly as compared to controls (FAB total score = 16.53 vs. 17.09, p .05). CCU had also a poor performance in the Motor Programming subtest (2.47 vs. 2.73, p .05). Conclusion: This study examined effects of cannabis in executive functioning and showed evidence that the FAB is sensitive to detect EF deficits in early abstinent chronic cannabis users. Clinical significance of these findings remains to be investigated in further longitudinal studies. FAB may be useful as a screening instrument to evaluate the necessity for a complete neuropsychological assessment in this population.

A Simple Technique for Large Tumor Removal During Laparoscopic Liver Resection

Experience with laparoscopic liver resections has increased in recent years, and so have the number of patients operated on by minimally invasive techniques. Specimen extraction is an important step of laparoscopic liver resection. The size of the specimen, is Usually a limitation for the use Of laparoscopy. The aim of this paper is to describe a new technique combining Pfannenstiel suprapubic incision and obstetric forceps to remove a large specimen from laparoscopic liver resections. The present technique allows an expeditious extraction of intact specimens, even huge ones, through a standard suprapubic Pfannenstiel incision. This technique has additional functional and cosmetic advantages over other techniques of specimen retrieval. We believe that the described technique is feasible, can be easily and rapidly performed, and facilitates laparoscopic liver resection by reducing the technical difficulties for specimen removal and may also be used in other abdominal laparoscopic interventions that deal with large surgical specimens.

Evaluation of erythrocyte dysmorphism by light microscopy with lowering of the condenser lens: A simple and efficient method

Aim: To demonstrate that the evaluation of erythrocyte dysmorphism by light microscopy with lowering of the condenser lens (LMLC) is useful to identify patients with a haematuria of glomerular or non-glomerular origin. Methods: A comparative double-blind study between phase contrast microscopy (PCM) and LMLC is reported to evaluate the efficacy of these techniques. Urine samples of 39 patients followed up for 9 months were analyzed, and classified as glomerular and non-glomerular haematuria. The different microscopic techniques were compared using receiver-operator curve (ROC) analysis and area under curve (AUC). Reproducibility was assessed by coefficient of variation (CV). Results: Specific cut-offs were set for each method according to their best rate of specificity and sensitivity as follows: 30% for phase contrast microscopy and 40% for standard LMLC, reaching in the first method the rate of 95% and 100% of sensitivity and specificity, respectively, and in the second method the rate of 90% and 100% of sensitivity and specificity, respectively. In ROC analysis, AUC for PCM was 0.99 and AUC for LMLC was 0.96. The CV was very similar in glomerular haematuria group for PCM (35%) and LMLC (35.3%). Conclusion: LMLC proved to be effective in contributing to the direction of investigation of haematuria, toward the nephrological or urological side. This method can substitute PCM when this equipment is not available.

A Simple and Efficient Device for Demonstrating Cross-sectional Anatomy of the Head

Described in this article is a novel device that facilitates study of the cross-sectional anatomy of the human head. In designing our device, we aimed to protect sections of the head from the destructive action of handling during anatomy laboratory while also ensuring excellent visualization of the anatomic structures. We used an electric saw to create 15-mm sections of three cadaver heads in the three traditional anatomic planes and inserted each section into a thin, perforated display box made of transparent acrylic material. The thin display boxes with head sections are kept in anatomical order in a larger transparent acrylic storage box containing formaldehyde solution, which preserves the specimens but also permits direct observation of the structures and their anatomic relationships to each other. This box-within-box design allows students to easily view sections of a head in its anatomical position as well as to examine internal structures by manipulating individual display boxes without altering the integrity of the preparations. This methodology for demonstrating cross-section anatomy allows efficient use of cadaveric material and technician time while also giving learners the best possible handling and visualization of complex anatomic structures. Our approach to teaching cross-sectional anatomy of the head can be applied to any part of human body, and the value of our device design will only increase as more complicated understandings of cross-sectional anatomy are required by advances and proliferation of imaging technology. Anat Sci Educ 3: 141-143, 2010. (C) 2010 American Association of Anatomists.

Influence of Photoactivation Protocol and Light Guide Distance on Conversion and Microleakage of Composite Restorations

Purpose: To evaluate the effect of light guide distance and the different photoactivation methods on the degree of conversion (DC) and microleakage of a composite. Methods and Materials: Three photoactivation protocols (600mW/cm(2) x 40 seconds; 400 mW/cm(2) x 60 seconds or 200 mW/cm(2) x 20 seconds, followed by 500 mW/cm(2) X 40 seconds) and three distances from the light source (0, 3 or 7 mm) were tested. Cylindrical specimens (5 nun diameter; 2 mm tall; n=3) were prepared for the DC test (FT-Raman). Class V cavities were made in 90 bovine incisors to conduct the microleakage test. The specimens were conditioned for 15 seconds with phosphoric acid (37%), followed by application of the adhesive system Prime & Bond NT (Dentsply/Caulk). The preparations were restored in bulk. The specimens were stored for 24 hours in distilled water (37 degrees C) before being submitted to the silvernitrate microleakage protocol. The restorations were sectioned and analyzed under 25x magnification. Results: Statistical analyses (two-way ANOVAs and Tukey test, alpha=0.05) found significance only for the factor distance (p=0.015) at the top of the composite for the DC test. Conversion was statistically lower for the 7 mm groups compared to the 0 and 3 mm groups, which were equivalent to each other. At the bottom of the specimens, none of the factors or interactions was significant (p<0.05). The Kruskal-Wallis test showed that, in general, the soft-start method led to lower microleakage scores when compared to the continuous modes, mainly when associated with a distancing of 7 mm (p<0.01). With the exception of specimens irradiated with 400mW/cm(2) that did not demonstrate variations on scores for the distances tested, higher microleakage was observed for shorter distances from the light source. Conclusions: Soft-start methods may reduce microleakage when the light guide distancing provides a low level of irradiance, which also causes a discrete reduction in the DC.

A simple, inexpensive and easily reproducible model of spinal cord injury in mice: Morphological and functional assessment

Spinal cord injury (SCI) causes motor and sensory deficits that impair functional performance, and significantly impacts life expectancy and quality. Animal models provide a good opportunity to test therapeutic strategies in vivo. C57BL/6 mice were subjected to laminectomy at T9 and compression with a vascular clip (30 g force, 1 min). Two groups were analyzed: injured group (SCI, n = 33) and laminectomy only (Sham, n = 15). Locomotor behavior (Basso mouse scale-BMS and global mobility) was assessed weekly. Morphological analyses were performed by LM and EM. The Sham group did not show any morphofunctional alteration. All SCI animals showed flaccid paralysis 24 h after injury. with subsequent improvement. The BMS score of the SCI group improved until the intermediate phase (2.037 +/- 1.198): the Sham animals maintained the highest BMS score (8.981 +/- 0.056). p < 0.001 during the entire time. The locomotor speed was slower in the SCI animals (5.581 +/- 0.871) than in the Sham animals (15.80 +/- 1.166), p < 0.001. Morphological analysis of the SCI group showed, in the acute phase, edema, hemorrhage, multiple cavities, fiber degeneration, cell death and demyelination. In the chronic phase we observed glial scarring, neuron death, and remyelination of spared axons by oligodendrocytes and Schwann cells. In conclusion, we established a simple, reliable, and inexpensive clip compression model in mice, with functional and morphological reproducibility and good validity. The availability of producing reliable injuries with appropriate outcome measures represents great potential for studies involving cellular mechanisms of primary injury and repair after traumatic SCI. (C) 2008 Elsevier B.V. All rights reserved.

Influence of the irradiation distance and the use of cooling to increase enamel-acid resistance with Er:YAG laser

Objectives: The aim of this study was to assess the influence of irradiation distance and the use of cooling in the Er:YAG laser efficacy in preventing enamel demineralization. Methods: 84 enamel blocks were randomly assigned to seven groups (n = 12): G1: control group - no treatment, G2-G7: experimental groups treated with Er:YAG laser (80 mJ/2 Hz) at different irradiation distances with or without cooling: G2: 4 mm/2 mL; G3: 4 mm/no cooling; G4: 8 mm/2 mL; G5: 8 mm/no cooling; G6: 16 mm/2 mL; G7: 16 mm/no cooling. The samples were submitted to an in vitro pH cycles for 14 days. Next, the specimens were sectioned in sections of 80-100 mu m in thickness and the demineralization patterns of prepared slices were assessed using a polarized light microscope. Three samples from each group were analyzed with scanning electronic microscopy. Analysis of variance and the Fisher test were performed for the statistical analysis of the data obtained from the caries-lesion-depth measurements (CLDM) (alpha = 5%). Results: The control group (CLDM = 0.67 mm) was statistically different from group 2 (CLDM = 0.42 mm), which presented a smaller lesion depth, and group 6 (0.91 mm), which presented a greater lesion depth. The results of groups 3 (CLDM = 0.74 mm), 4 (CLDM = 0.70 mm), 5 (CLDM = 0.67 mm) and 7 (CLDM = 0.89 mm) presented statistical similarity. The scanning electronic microscopy analysis showed ablation areas in the samples from groups 4, 5, 6 and 7, and a slightly demineralized area in group 2. Conclusions: It was possible to conclude that Er:YAG laser was efficient in preventing enamel demineralization at a 4-mm irradiation distance using cooling. (C) 2010 Elsevier Ltd. All rights reserved.