Using Manifold Learning for Atlas Selection in Multi-Atlas Segmentation

Multi-atlas segmentation has been widely used to segment various anatomical structures. The success of this technique partly relies on the selection of atlases that are best mapped to a new target image after registration. Recently, manifold learning has been proposed as a method for atlas selection. Each manifold learning technique seeks to optimize a unique objective function. Therefore, different techniques produce different embeddings even when applied to the same data set. Previous studies used a single technique in their method and gave no reason for the choice of the manifold learning technique employed nor the theoretical grounds for the choice of the manifold parameters. In this study, we compare side-by-side the results given by 3 manifold learning techniques (Isomap, Laplacian Eigenmaps and Locally Linear Embedding) on the same data set. We assess the ability of those 3 different techniques to select the best atlases to combine in the framework of multi-atlas segmentation. First, a leave-one-out experiment is used to optimize our method on a set of 110 manually segmented atlases of hippocampi and find the manifold learning technique and associated manifold parameters that give the best segmentation accuracy. Then, the optimal parameters are used to automatically segment 30 subjects from the Alzheimer’s Disease Neuroimaging Initiative (ADNI). For our dataset, the selection of atlases with Locally Linear Embedding gives the best results. Our findings show that selection of atlases with manifold learning leads to segmentation accuracy close to or significantly higher than the state-of-the-art method and that accuracy can be increased by fine tuning the manifold learning process.     

Rib lesions in skeletons from early neolithic sites in Central Germany: On the trail of tuberculosis at the onset of agriculture

As an infectious disease, tuberculosis (TB) is one of the major causes of death worldwide. Paleopathological and paleomicrobiological studies indicate a long standing association of the causative agent Mycobacterium tuberculosis and its human host. Since the occurrence and the epidemic spread of this pathogen seem to be closely linked to social and biological factors, it is of particular interest to understand better the role of TB during periods of social and nutritional change such as the Neolithic. In this study, 118 individuals from three sites in Saxony‐Anhalt (Germany) dating to the Linear Pottery Culture (5400–4800 BC) were examined macroscopically to identify TB related bone lesions. In two individuals, Pott's disease was detected. In addition, periosteal reactions of varying degrees and frequency were observed mainly along the neck of the ribs in 6.5% (2/31) of subadults and 35.1% (20/57) of adults, with one site standing out markedly. Rib lesions, however, are not specific indicators of TB as they can also be caused by other diseases; so additional investigations were undertaken using histology and micro‐CT scans to say more about the disease process. Supplementary molecular analyses indicate the presence of pathogens belonging to the Mycobacterium tuberculosis complex in individuals of all sites. Furthermore, we discuss the occurrence and spread of TB during the Neolithic with regard to nutritional aspects and possible risks of infection. The data presented provide important insights into the health status of Early Neolithic populations in Central Germany. Am J Phys Anthropol, 2012. © 2012 Wiley Periodicals, Inc.     

Forecasting Temporal Dynamics of Cutaneous Leishmaniasis in Northeast Brazil

Introduction

Cutaneous leishmaniasis (CL) is a vector-borne disease of increasing importance in northeastern Brazil. It is known that sandflies, which spread the causative parasites, have weather-dependent population dynamics. Routinely-gathered weather data may be useful for anticipating disease risk and planning interventions.

Methodology/Principal Findings

We fit time series models using meteorological covariates to predict CL cases in a rural region of Bahía, Brazil from 1994 to 2004. We used the models to forecast CL cases for the period 2005 to 2008. Models accounting for meteorological predictors reduced mean squared error in one, two, and three month-ahead forecasts by up to 16% relative to forecasts from a null model accounting only for temporal autocorrelation.

Significance

These outcomes suggest CL risk in northeastern Brazil might be partially dependent on weather. Responses to forecasted CL epidemics may include bolstering clinical capacity and disease surveillance in at-risk areas. Ecological mechanisms by which weather influences CL risk merit future research attention as public health intervention targets.     

Small-scale field evaluation of the efficacy and residual effect of Fludora® Fusion (mixture of clothianidin and deltamethrin) against susceptible and resistant Anopheles gambiae populations from Benin,West Africa

The emergence of mosquitoes that can avoid indoor-deployed interventions, such as treated bed nets and indoor residual spraying, threatens the mainstay of malaria control in Zambia. Furthermore, the requirement for high coverage of these tools poses operational challenges. Spatial repellents are being assessed to supplement these vector control tools, but limitations exist in the residual effect of the repellent and the need for external power or heat for diffusion of the volatiles. A semi-field evaluation of a novel controlled release spatial repellent device (CRD) was conducted in Macha, Zambia. These devices emanate metofluthrin with no need for external power. Devices were deployed in huts within the semi-field system (SFS). Female Anopheles gambiae sensu stricto released within the SFS were trapped overnight by light traps and collected by aspiration the next morning inside and outside of huts to determine the extent of mosquito repellency and the impact on host-seeking and survival. Experiments studied the impact of number of devices as well as the presence of hut occupants. The study was complemented with numerical methods based on computational fluid dynamics to simulate spatial distribution of metofluthrin. Presence of CRDs was associated with significant reductions in indoor counts of mosquitoes, regardless of whether huts were occupied or not. Repellency ranged from 15 to 60% compared to huts with no devices. Reducing the number of devices from 16 to 4 had little impact on repellency. When huts were occupied, indoor mosquito host-seeking was higher in the presence of CRDs, whilst survival was significantly reduced. This study demonstrated that deployment of as few as four CRDs within a hut was associated with reduced indoor mosquito densities. As would be expected, presence of occupants within huts, resulted in greater indoor catches (both with and without devices). The increased indoor mosquito host-seeking and mortality in huts when devices were present may be explained by the excito-repellency activity of metofluthrin. These semi-field experiments provide preliminary data on the utility of CRD spatial repellents to reduce indoor densities of An. gambiae mosquitoes. Studies will further investigate the impact of CRDs on mosquito behaviour as well as epidemiological protective efficacy.     

The PlA1/A2 Polymorphism of Glycoprotein IIIa as a Risk Factor for Myocardial Infarction: A Meta-Analysis

Background

The PlA2 polymorphism of glycoprotein IIIa (GPIIIa) has been previously identified as being associated with myocardial infarction (MI), but whether this represents a true association is entirely unclear due to differences in findings from different studies. We performed a meta-analysis to evaluate whether this polymorphism is a risk factor for MI.

Methods

Electronic databases (MEDLINE and EMBASE) were searched for all articles evaluating genetic polymorphisms of GPIIIa. For studies where acute coronary events were recorded in association with genetic analysis, pooled odds ratios (ORs) were calculated using fixed-effects and random-effects models. The primary outcome measure was MI, and a secondary analysis was also performed for acute coronary syndromes (ACS) more generally.

Findings

57 studies were eligible for statistical analysis and included 17,911 cases and 24,584 controls. Carriage of the PlA2 allele was significantly associated with MI (n = 40,692; OR 1.077, 95% CI 1.024–1.132; p = 0.004) but with significant publication bias (p = 0.040). The degree of association with MI increased with decreasing age of subjects (≤45 years old: n = 9,547; OR 1.205, 95% CI 1.067–1.360; p = 0.003) and with adjustment of data for conventional cardiovascular risk factors (n = 12,001; OR 1.240, 95% CI 1.117–1.376; p<0.001). There was a low probability of publication bias for these subgroup analyses (all p<0.05).

Conclusions

The presence of significant publication bias makes it unclear whether the association between carriage of the PlA2 allele and MI is true for the total population studied. However for younger subjects, the relative absence of conventional cardiovascular risk factors results in a significant association between carriage of the PlA2 allele and MI.