Supporting content-based image retrieval and computer-aided diagnosis systems with association rule-based techniques

In this work, we take advantage of association rule mining to support two types of medical systems: the Content-based Image Retrieval (CBIR) systems and the Computer-Aided Diagnosis (CAD) systems. For content-based retrieval, association rules are employed to reduce the dimensionality of the feature vectors that represent the images and to improve the precision of the similarity queries. We refer to the association rule-based method to improve CBIR systems proposed here as Feature selection through Association Rules (FAR). To improve CAD systems, we propose the Image Diagnosis Enhancement through Association rules (IDEA) method. Association rules are employed to suggest a second opinion to the radiologist or a preliminary diagnosis of a new image. A second opinion automatically obtained can either accelerate the process of diagnosing or to strengthen a hypothesis, increasing the probability of a prescribed treatment be successful. Two new algorithms are proposed to support the IDEA method: to pre-process low-level features and to propose a preliminary diagnosis based on association rules. We performed several experiments to validate the proposed methods. The results indicate that association rules can be successfully applied to improve CBIR and CAD systems, empowering the arsenal of techniques to support medical image analysis in medical systems. (C) 2009 Elsevier B.V. All rights reserved.

Evidence for a network transcriptional control of promiscuous gene expression in medullary thymic epithelial cells

The expression of peripheral tissue antigens (PTAs) in the thymus by medullary thymic epithelial cells (mTECs) is essential for the central self-tolerance in the generation of the T cell repertoire. Due to heterogeneity of autoantigen representation, this phenomenon has been termed promiscuous gene expression (PGE), in which the autoimmune regulator (Aire) gene plays a key role as a transcription factor in part of these genes. Here we used a microarray strategy to access PGE in cultured murine CD80(+) 3.10 mTEC line. Hierarchical clustering of the data allowed observation that PTA genes were differentially expressed being possible to found their respective induced or repressed mRNAs. To further investigate the control of PGE, we tested the hypothesis that genes involved in this phenomenon might also be modulated by transcriptional network. We then reconstructed such network based on the microarray expression data, featuring the guanylate cyclase 2d (Gucy2d) gene as a main node. In such condition, we established 167 positive and negative interactions with downstream PTA genes. Silencing Aire by RNA interference, Gucy2d while down regulated established a larger number (355) of interactions with PTA genes. T- and G-boxes corresponding to AIRE protein binding sites located upstream to ATG codon of Gucy2d supports this effect. These findings provide evidence that Aire plays a role in association with Gucy2d, which is connected to Several PTA genes and establishes a cascade-like transcriptional control of promiscuous gene expression in mTEC cells. (C) 2009 Elsevier Ltd. All rights reserved.

An association rule-based method to support medical image diagnosis with efficiency

In this paper, we propose a method based on association rule-mining to enhance the diagnosis of medical images (mammograms). It combines low-level features automatically extracted from images and high-level knowledge from specialists to search for patterns. Our method analyzes medical images and automatically generates suggestions of diagnoses employing mining of association rules. The suggestions of diagnosis are used to accelerate the image analysis performed by specialists as well as to provide them an alternative to work on. The proposed method uses two new algorithms, PreSAGe and HiCARe. The PreSAGe algorithm combines, in a single step, feature selection and discretization, and reduces the mining complexity. Experiments performed on PreSAGe show that this algorithm is highly suitable to perform feature selection and discretization in medical images. HiCARe is a new associative classifier. The HiCARe algorithm has an important property that makes it unique: it assigns multiple keywords per image to suggest a diagnosis with high values of accuracy. Our method was applied to real datasets, and the results show high sensitivity (up to 95%) and accuracy (up to 92%), allowing us to claim that the use of association rules is a powerful means to assist in the diagnosing task.

Linkage to chromosome 2q36.1 in autosomal dominant Dandy-Walker malformation with occipital cephalocele and evidence for genetic heterogeneity

We previously reported a Vietnamese-American family with isolated autosomal dominant occipital cephalocele. Upon further neuroimaging studies, we have recharacterized this condition as autosomal dominant Dandy-Walker with occipital cephalocele (ADDWOC). A similar ADDWOC family from Brazil was also recently described. To determine the genetic etiology of ADDWOC, we performed genome-wide linkage analysis on members of the Vietnamese-American and Brazilian pedigrees. Linkage analysis of the Vietnamese-American family identified the ADDWOC causative locus on chromosome 2q36.1 with a multipoint parametric LOD score of 3.3, while haplotype analysis refined the locus to 1.1 Mb. Sequencing of the five known genes in this locus did not identify any protein-altering mutations. However, a terminal deletion of chromosome 2 in a patient with an isolated case of Dandy-Walker malformation also encompassed the 2q36.1 chromosomal region. The Brazilian pedigree did not show linkage to this 2q36.1 region. Taken together, these results demonstrate a locus for ADDWOC on 2q36.1 and also suggest locus heterogeneity for ADDWOC.

Anti-mullerian hormone is the best predictor of poor response in ICSI cycles of patients with endometriosis

Purpose: To correlate ovarian reserve (OR) markers with response in assisted reproduction techniques (ART) and determine their ability to predict poor response among patients with endometriosis (EDT). Methods: We evaluated ART cycles of 27 women with EDT and 50 with exclusive male factor. Basal follicle stimulating hormone (FSH) and anti-mullerian hormone (AMH) levels were determined. Ovarian response to gonadotropin stimulation was assessed and correlation coefficients calculated between the variables and reserve markers. Areas under the curve (AUC) determined ability of tests to predict poor response. Results: AMH was significantly correlated with response in both groups and it was the only marker with significant discriminative capacity to predict poor response among EDT (AUC = 0.842; 95% CI: 0.651-0.952) and control group (AUC = 0.869; 95% CI: 0.743-0.947). Conclusion: Infertile patients with endometriosis can benefit from the pre-therapeutic assessment of OR markers. However, regardless of disease presence, only AMH predicts poor response to stimulus.

What is the best equation to estimate the basal energy expenditure of climacteric women?

Objectives The methods currently available for the measurement of energy expenditure in patients, such as indirect calorimetry and double-labelled water, are expensive and are limited in Brazil to research projects. Thus, equations for the prediction of resting metabolic rate appear to be a viable alternative for clinical practice. However, there are no specific equations for the Brazilian population and few studies have been conducted on Brazilian women in the climacteric period using existing and commonly applied equations. On this basis, the objective of the present study was to investigate the concordance between the predictive equations most frequently used and indirect calorimetry for the measurement of resting metabolic rate. Methods We calculated the St. Laurent concordance correlation coefficient between the equations and resting metabolic rate calculated by indirect calorimetry in 46 climacteric women. Results The equation showing the best concordance was that of the FAO/WHO/UNU formula (0.63), which proved to be better than the Harris & Benedict equation (0.55) for the sample studied. Conclusions On the basis of the results of the present study, we conclude that the FAO/WHO/UNU formula can be used to predict better the resting metabolic rate of climacteric women. Further studies using more homogeneous and larger samples are needed to permit the use of the FAO/WHO/UNU formula for this population group with greater accuracy.

Effects of beta-carboline harmine on behavioral and physiological parameters observed in the chronic mild stress model: Further evidence of antidepressant properties

The chronic mild stress (CMS) model has been used as an animal model of depression which induces anhedonic behavior in rodents. The present study was aimed to evaluate the behavioral and physiological effects of administration of P-carboline harmine in rats exposed to CMS Procedure. To this aim, after 40 days of exposure to CMS procedure, rats were treated with harmine (15 mg/kg/day) for 7 days. In this study, sweet food consumption, adrenal gland weight, adrenocorticotrophin hormone (ACTH) levels, and hippocampal brain-derived-neurotrophic factor (BDNF) protein levels were assessed. Our findings demonstrated that chronic stressful situations induced anhedonia, hypertrophy of adrenal gland weight, increase ACTH circulating levels in rats and increase BDNF protein levels. Interestingly, treatment with harmine reversed anhedonia, the increase of adrenal gland weight, normalized ACTH circulating levels and BDNF protein levels. Finally, these findings further support the hypothesis that harmine could be a new pharmacological tool for the treatment of depression. (C) 2009 Elsevier Inc. All rights reserved.

Evidence of Renal Infection in Fatal Cases of 2009 Pandemic Influenza A (HI NI)

The 2009 pandemic influenza A (H1N1) caused significant morbidity and mortality. Acute lung injury is the hallmark of the disease, but multiple organ system dysfunction can develop and lead to death. Therefore, we sought to investigate whether there was postmortem evidence of H1N1 presence and virus-induced organ injury in autopsy specimens. Five cases in which patients died of influenza A (H1N1) virus infection were studied. The lungs of all patients showed macroscopic and microscopic findings already described for H1N1 (consolidation, edema, hemorrhage, alveolar damage, hyaline membrane, and inflammation), and H1N1 viruses were present in alveolar cells in immunochemical studies. Acute tubular necrosis was present in all cases, but there was no evidence of direct virus-induced kidney injury. Nevertheless, H1N1 viruses were found in the cytoplasm of glomerular macrophages in the kidneys of 4 patients. Therefore, our data provide strong evidence that H1N1 presence is not restricted to the lungs.