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.

Variation in the type and frequency of postoperative invasive Staphylococcus aureus infections according to type of surgical procedure.

OBJECTIVE: To determine the epidemiological characteristics of postoperative invasive Staphylococcus aureus infection following 4 types of major surgical procedures.design. Retrospective cohort study. SETTING: Eleven hospitals (9 community hospitals and 2 tertiary care hospitals) in North Carolina and Virginia. PATIENTS: Adults undergoing orthopedic, neurosurgical, cardiothoracic, and plastic surgical procedures. METHODS: We used previously validated, prospectively collected surgical surveillance data for surgical site infection and microbiological data for bloodstream infection. The study period was 2003 through 2006. We defined invasive S. aureus infection as either nonsuperficial incisional surgical site infection or bloodstream infection. Nonparametric bootstrapping was used to generate 95% confidence intervals (CIs). P values were generated using the Pearson chi2 test, Student t test, or Wilcoxon rank-sum test, as appropriate. RESULTS: In total, 81,267 patients underwent 96,455 procedures during the study period. The overall incidence of invasive S. aureus infection was 0.47 infections per 100 procedures (95% CI, 0.43-0.52); 227 (51%) of 446 infections were due to methicillin-resistant S.aureus. Invasive S. aureus infection was more common after cardiothoracic procedures (incidence, 0.79 infections per 100 procedures [95%CI, 0.62-0.97]) than after orthopedic procedures (0.37 infections per 100 procedures [95% CI, 0.32-0.42]), neurosurgical procedures (0.62 infections per 100 procedures [95% CI, 0.53-0.72]), or plastic surgical procedures (0.32 infections per 100 procedures [95% CI, 0.17-0.47]) (P < .001). Similarly, S. aureus bloodstream infection was most common after cardiothoracic procedures (incidence, 0.57 infections per 100 procedures [95% CI, 0.43-0.72]; P < .001, compared with other procedure types), comprising almost three-quarters of the invasive S. aureus infections after these procedures. The highest rate of surgical site infection was observed after neurosurgical procedures (incidence, 0.50 infections per 100 procedures [95% CI, 0.42-0.59]; P < .001, compared with other procedure types), comprising 80% of invasive S.aureus infections after these procedures. CONCLUSION: The frequency and type of postoperative invasive S. aureus infection varied significantly across procedure types. The highest risk procedures, such as cardiothoracic procedures, should be targeted for ongoing preventative interventions.

Genetic signatures in the envelope glycoproteins of HIV-1 that associate with broadly neutralizing antibodies.

A steady increase in knowledge of the molecular and antigenic structure of the gp120 and gp41 HIV-1 envelope glycoproteins (Env) is yielding important new insights for vaccine design, but it has been difficult to translate this information to an immunogen that elicits broadly neutralizing antibodies. To help bridge this gap, we used phylogenetically corrected statistical methods to identify amino acid signature patterns in Envs derived from people who have made potently neutralizing antibodies, with the hypothesis that these Envs may share common features that would be useful for incorporation in a vaccine immunogen. Before attempting this, essentially as a control, we explored the utility of our computational methods for defining signatures of complex neutralization phenotypes by analyzing Env sequences from 251 clonal viruses that were differentially sensitive to neutralization by the well-characterized gp120-specific monoclonal antibody, b12. We identified ten b12-neutralization signatures, including seven either in the b12-binding surface of gp120 or in the V2 region of gp120 that have been previously shown to impact b12 sensitivity. A simple algorithm based on the b12 signature pattern was predictive of b12 sensitivity/resistance in an additional blinded panel of 57 viruses. Upon obtaining these reassuring outcomes, we went on to apply these same computational methods to define signature patterns in Env from HIV-1 infected individuals who had potent, broadly neutralizing responses. We analyzed a checkerboard-style neutralization dataset with sera from 69 HIV-1-infected individuals tested against a panel of 25 different Envs. Distinct clusters of sera with high and low neutralization potencies were identified. Six signature positions in Env sequences obtained from the 69 samples were found to be strongly associated with either the high or low potency responses. Five sites were in the CD4-induced coreceptor binding site of gp120, suggesting an important role for this region in the elicitation of broadly neutralizing antibody responses against HIV-1.

Musculoskeletal symptoms among female garment factory workers in Sri Lanka.

OBJECTIVES: To assess the prevalence of musculoskeletal symptoms and their association with sociodemographic risk factors among female garment factory workers in Sri Lanka. METHODS: 1058 randomly selected female garment factory workers employed in the free trade zone of Kogalla, Sri Lanka were recruited to complete two interviewer-administered questionnaires assessing musculoskeletal symptoms and health behaviors. DISCUSSION: Musculoskeletal complaints among female garment workers in the FTZ of Kogalla are less common than expected. Sociocultural factors may have resulted in underreporting and similarly contribute to the low rates of healthcare utilization by these women. RESULTS: 164 (15.5%) of workers reported musculoskeletal symptoms occurring more than 3 times or lasting a week or more during the previous 12-month period. Back (57.3%) and knee (31.7%) were the most common sites of pain. Although most symptomatic women reported that their problems interfered with work and leisure activities, very few missed work as a result of their pain. Prevalence correlated positively with increased age and industry tenure of less than 12 months. Job type, body mass index, and education were not significant predictors of musculoskeletal symptoms.

Four-decade responses of soil trace elements to an aggrading old-field forest: B, Mn, Zn, Cu, and Fe.

In the ancient and acidic Ultisol soils of the Southern Piedmont, USA, we studied changes in trace element biogeochemistry over four decades, a period during which formerly cultivated cotton fields were planted with pine seedlings that grew into mature forest stands. In 16 permanent plots, we estimated 40-year accumulations of trace elements in forest biomass and O horizons (between 1957 and 1997), and changes in bioavailable soil fractions indexed by extractions of 0.05 mol/L HCl and 0.2 mol/L acid ammonium oxalate (AAO). Element accumulations in 40-year tree biomass plus O horizons totaled 0.9, 2.9, 4.8, 49.6, and 501.3 kg/ha for Cu, B, Zn, Mn, and Fe, respectively. In response to this forest development, samples of the upper 0.6-m of mineral soil archived in 1962 and 1997 followed one of three patterns. (1) Extractable B and Mn were significantly depleted, by -4.1 and -57.7 kg/ha with AAO, depletions comparable to accumulations in biomass plus O horizons, 2.9 and 49.6 kg/ha, respectively. Tree uptake of B and Mn from mineral soil greatly outpaced resupplies from atmospheric deposition, mineral weathering, and deep-root uptake. (2) Extractable Zn and Cu changed little during forest growth, indicating that nutrient resupplies kept pace with accumulations by the aggrading forest. (3) Oxalate-extractable Fe increased substantially during forest growth, by 275.8 kg/ha, about 10-fold more than accumulations in tree biomass (28.7 kg/ha). The large increases in AAO-extractable Fe in surficial 0.35-m mineral soils were accompanied by substantial accretions of Fe in the forest's O horizon, by 473 kg/ha, amounts that dwarfed inputs via litterfall and canopy throughfall, indicating that forest Fe cycling is qualitatively different from that of other macro- and micronutrients. Bioturbation of surficial forest soil layers cannot account for these fractions and transformations of Fe, and we hypothesize that the secondary forest's large inputs of organic additions over four decades has fundamentally altered soil Fe oxides, potentially altering the bioavailability and retention of macro- and micronutrients, contaminants, and organic matter itself. The wide range of responses among the ecosystem's trace elements illustrates the great dynamics of the soil system over time scales of decades.

Dynamic neural networks supporting memory retrieval.

How do separate neural networks interact to support complex cognitive processes such as remembrance of the personal past? Autobiographical memory (AM) retrieval recruits a consistent pattern of activation that potentially comprises multiple neural networks. However, it is unclear how such large-scale neural networks interact and are modulated by properties of the memory retrieval process. In the present functional MRI (fMRI) study, we combined independent component analysis (ICA) and dynamic causal modeling (DCM) to understand the neural networks supporting AM retrieval. ICA revealed four task-related components consistent with the previous literature: 1) medial prefrontal cortex (PFC) network, associated with self-referential processes, 2) medial temporal lobe (MTL) network, associated with memory, 3) frontoparietal network, associated with strategic search, and 4) cingulooperculum network, associated with goal maintenance. DCM analysis revealed that the medial PFC network drove activation within the system, consistent with the importance of this network to AM retrieval. Additionally, memory accessibility and recollection uniquely altered connectivity between these neural networks. Recollection modulated the influence of the medial PFC on the MTL network during elaboration, suggesting that greater connectivity among subsystems of the default network supports greater re-experience. In contrast, memory accessibility modulated the influence of frontoparietal and MTL networks on the medial PFC network, suggesting that ease of retrieval involves greater fluency among the multiple networks contributing to AM. These results show the integration between neural networks supporting AM retrieval and the modulation of network connectivity by behavior.