Optimized approach to decision fusion of heterogeneous data for breast cancer diagnosis.

As more diagnostic testing options become available to physicians, it becomes more difficult to combine various types of medical information together in order to optimize the overall diagnosis. To improve diagnostic performance, here we introduce an approach to optimize a decision-fusion technique to combine heterogeneous information, such as from different modalities, feature categories, or institutions. For classifier comparison we used two performance metrics: The receiving operator characteristic (ROC) area under the curve [area under the ROC curve (AUC)] and the normalized partial area under the curve (pAUC). This study used four classifiers: Linear discriminant analysis (LDA), artificial neural network (ANN), and two variants of our decision-fusion technique, AUC-optimized (DF-A) and pAUC-optimized (DF-P) decision fusion. We applied each of these classifiers with 100-fold cross-validation to two heterogeneous breast cancer data sets: One of mass lesion features and a much more challenging one of microcalcification lesion features. For the calcification data set, DF-A outperformed the other classifiers in terms of AUC (p < 0.02) and achieved AUC=0.85 +/- 0.01. The DF-P surpassed the other classifiers in terms of pAUC (p < 0.01) and reached pAUC=0.38 +/- 0.02. For the mass data set, DF-A outperformed both the ANN and the LDA (p < 0.04) and achieved AUC=0.94 +/- 0.01. Although for this data set there were no statistically significant differences among the classifiers' pAUC values (pAUC=0.57 +/- 0.07 to 0.67 +/- 0.05, p > 0.10), the DF-P did significantly improve specificity versus the LDA at both 98% and 100% sensitivity (p < 0.04). In conclusion, decision fusion directly optimized clinically significant performance measures, such as AUC and pAUC, and sometimes outperformed two well-known machine-learning techniques when applied to two different breast cancer data sets.

Auriculotherapy for pain management: a systematic review and meta-analysis of randomized controlled trials.

OBJECTIVES: Side-effects of standard pain medications can limit their use. Therefore, nonpharmacologic pain relief techniques such as auriculotherapy may play an important role in pain management. Our aim was to conduct a systematic review and meta-analysis of studies evaluating auriculotherapy for pain management. DESIGN: MEDLINE,(®) ISI Web of Science, CINAHL, AMED, and Cochrane Library were searched through December 2008. Randomized trials comparing auriculotherapy to sham, placebo, or standard-of-care control were included that measured outcomes of pain or medication use and were published in English. Two (2) reviewers independently assessed trial eligibility, quality, and abstracted data to a standardized form. Standardized mean differences (SMD) were calculated for studies using a pain score or analgesic requirement as a primary outcome. RESULTS: Seventeen (17) studies met inclusion criteria (8 perioperative, 4 acute, and 5 chronic pain). Auriculotherapy was superior to controls for studies evaluating pain intensity (SMD, 1.56 [95% confidence interval (CI): 0.85, 2.26]; 8 studies). For perioperative pain, auriculotherapy reduced analgesic use (SMD, 0.54 [95% CI: 0.30, 0.77]; 5 studies). For acute pain and chronic pain, auriculotherapy reduced pain intensity (SMD for acute pain, 1.35 [95% CI: 0.08, 2.64], 2 studies; SMD for chronic pain, 1.84 [95% CI: 0.60, 3.07], 5 studies). Removal of poor quality studies did not alter the conclusions. Significant heterogeneity existed among studies of acute and chronic pain, but not perioperative pain. CONCLUSIONS: Auriculotherapy may be effective for the treatment of a variety of types of pain, especially postoperative pain. However, a more accurate estimate of the effect will require further large, well-designed trials.

Relationships between polybrominated diphenyl ether concentrations in house dust and serum.

Polybrominated diphenyl ethers (PBDEs) have been measured in the home environment and in humans, but studies linking environmental levels to body burdens are limited. This study examines the relationship between PBDE concentrations in house dust and serum from adults residing in these homes. We measured PBDE concentrations in house dust from 50 homes and in serum of male-female couples from 12 of the homes. Detection rates, dust-serum, and within-matrix correlations varied by PBDE congener. There was a strong correlation (r = 0.65-0.89, p < 0.05) between dust and serum concentrations of several predominant PBDE congeners (BDE 47, 99, and 100). Dust and serum levels of BDE 153 were not correlated (r < 0.01). The correlation of dust and serum levels of BDE 209 could not be evaluated due to low detection rates of BDE 209 in serum. Serum concentrations of the sum of BDE 47, 99, and 100 were also strongly correlated within couples (r = 0.85, p = 0.0005). This study provides evidence that house dust is a primary exposure pathway of PBDEs and supports the use of dust PBDE concentrations as a marker for exposure to PBDE congeners other than BDE 153.

Correlates of HIV testing among abused women in South Africa.

Gender-based violence increases a woman's risk for HIV but little is known about her decision to get tested. We interviewed 97 women seeking abuse-related services from a nongovernmental organization (NGO) in Johannesburg, South Africa. Forty-six women (47%) had been tested for HIV. Caring for children (odds ratio [OR] = 0.27, 95% confidence interval [CI] = [0.07, 1.00]) and conversing with partner about HIV (OR = 0.13, 95% CI = [0.02, 0.85]) decreased odds of testing. Stronger risk-reduction intentions (OR = 1.27, 95% CI = [1.01, 1.60]) and seeking help from police (OR = 5.51, 95% CI = [1.18, 25.76]) increased odds of testing. Providing safe access to integrated services and testing may increase testing in this population. Infection with HIV is highly prevalent in South Africa where an estimated 16.2% of adults between the ages of 15 and 49 have the virus. The necessary first step to stemming the spread of HIV and receiving life-saving treatment is learning one's HIV serostatus through testing. Many factors may contribute to someone's risk of HIV infection and many barriers may prevent testing. One factor that does both is gender-based violence.

Dental microwear in live, wild-trapped Alouatta palliata from Costa Rica.

One problem with dental microwear analyses of museum material is that investigators can never be sure of the diets of the animals in question. An obvious solution to this problem is to work with live animals. Recent work with laboratory primates has shown that high resolution dental impressions can be obtained from live animals. The purpose of this study was to use similar methods to begin to document rates and patterns of dental microwear for primates in the wild. Thirty-three Alouatta palliata were captured during the wet season at Hacienda La Pacifica near Canas, Costa Rica. Dental impressions were taken and epoxy casts of the teeth were prepared using the methods of Teaford and Oyen (1989a). Scanning electron micrographs were taken of the left mandibular second molars at magnifications of 200x and 500x. Lower magnification images were used to calculate rates of wear, and higher magnification images were used to measure the size and shape of microwear features. Results indicate that, while basic patterns of dental microwear are similar in museum samples and samples of live, wild-trapped animals of the same species, ecological differences between collection locales may lead to significant intraspecific differences in dental microwear. More importantly, rates of microwear provide the first direct evidence of differences in molar use between monkeys and humans.

Increased leptin levels correlate with thyroid autoantibodies in nonobese males.

OBJECTIVE: Leptin is an adipokine that regulates body weight and appetite. It is also an inflammatory cytokine that influences immune reactivity and autoimmunity. Leptin levels are increased in obesity and are higher in women than in men. We aimed to determine whether leptin levels, independent of sex and body mass index (BMI), are associated with thyroid autoimmunity. DESIGN: This study uses data from The Third National Health and Nutrition Examination Survey (NHANES III) to test the association of leptin and thyroid autoimmunity, independent of BMI. MEASUREMENTS: Thyroid-stimulating hormone, thyroxine, antithyroid peroxidase (TPO) antibodies and leptin levels were measured in 2902 men and 3280 women within the NHANES III population. BMI was calculated from height and weight. RESULTS: Women had significantly higher leptin levels and anti-TPO antibody titres than men. Correlation analyses demonstrated that leptin levels were associated with anti-TPO antibody levels in the total population, but when men and women were analysed separately, this association was lost. We then stratified men and women into obese (BMI > 30) or nonobese (BMI ≤ 30) subgroups and determined the association between leptin levels and anti-TPO antibody titres for each subgroup. Using regression analysis, we found that increased leptin levels correlated with thyroid autoantibodies in nonobese males, but not in obese males or in females. CONCLUSIONS: Leptin levels correlated with thyroid autoantibody titres in nonobese males. This association was not found in females. Sex and body habitus should therefore be considered in studying the role of leptin in other autoimmune conditions.