Exploring the knowledge contained in neuroimages: Statistical discriminant analysis and automatic segmentation of the most significant changes

Objective: The aim of this article is to propose an integrated framework for extracting and describing patterns of disorders from medical images using a combination of linear discriminant analysis and active contour models. Methods: A multivariate statistical methodology was first used to identify the most discriminating hyperplane separating two groups of images (from healthy controls and patients with schizophrenia) contained in the input data. After this, the present work makes explicit the differences found by the multivariate statistical method by subtracting the discriminant models of controls and patients, weighted by the pooled variance between the two groups. A variational level-set technique was used to segment clusters of these differences. We obtain a label of each anatomical change using the Talairach atlas. Results: In this work all the data was analysed simultaneously rather than assuming a priori regions of interest. As a consequence of this, by using active contour models, we were able to obtain regions of interest that were emergent from the data. The results were evaluated using, as gold standard, well-known facts about the neuroanatomical changes related to schizophrenia. Most of the items in the gold standard was covered in our result set. Conclusions: We argue that such investigation provides a suitable framework for characterising the high complexity of magnetic resonance images in schizophrenia as the results obtained indicate a high sensitivity rate with respect to the gold standard. (C) 2010 Elsevier B.V. All rights reserved.

A decision support system to facilitate management of patients with acute gastrointestinal bleeding

Objective: To develop a model to predict the bleeding source and identify the cohort amongst patients with acute gastrointestinal bleeding (GIB) who require urgent intervention, including endoscopy. Patients with acute GIB, an unpredictable event, are most commonly evaluated and managed by non-gastroenterologists. Rapid and consistently reliable risk stratification of patients with acute GIB for urgent endoscopy may potentially improve outcomes amongst such patients by targeting scarce health-care resources to those who need it the most. Design and methods: Using ICD-9 codes for acute GIB, 189 patients with acute GIB and all. available data variables required to develop and test models were identified from a hospital medical records database. Data on 122 patients was utilized for development of the model and on 67 patients utilized to perform comparative analysis of the models. Clinical data such as presenting signs and symptoms, demographic data, presence of co-morbidities, laboratory data and corresponding endoscopic diagnosis and outcomes were collected. Clinical data and endoscopic diagnosis collected for each patient was utilized to retrospectively ascertain optimal management for each patient. Clinical presentations and corresponding treatment was utilized as training examples. Eight mathematical models including artificial neural network (ANN), support vector machine (SVM), k-nearest neighbor, linear discriminant analysis (LDA), shrunken centroid (SC), random forest (RF), logistic regression, and boosting were trained and tested. The performance of these models was compared using standard statistical analysis and ROC curves. Results: Overall the random forest model best predicted the source, need for resuscitation, and disposition with accuracies of approximately 80% or higher (accuracy for endoscopy was greater than 75%). The area under ROC curve for RF was greater than 0.85, indicating excellent performance by the random forest model Conclusion: While most mathematical models are effective as a decision support system for evaluation and management of patients with acute GIB, in our testing, the RF model consistently demonstrated the best performance. Amongst patients presenting with acute GIB, mathematical models may facilitate the identification of the source of GIB, need for intervention and allow optimization of care and healthcare resource allocation; these however require further validation. (c) 2007 Elsevier B.V. All rights reserved.

Comparison of pelvic floor muscle strength between women undergoing vaginal delivery, cesarean section, and nulliparae using a perineometer and digital palpation

Objectives. To compare pelvic floor muscle (PFM) strength between women undergoing vaginal delivery, cesarean section, and nulliparae, investigating the factors associated with PFM strength, and observing the correlation between vaginal digital palpation and use of a perineometer. Methods. A cross-sectional study was conducted, including 31 women following vaginal delivery, 30 women following cesarean section, and 30 nulliparous women. PFM strength was measured by vaginal digital palpation and use of a perineometer. Multiple linear regression analysis with adjustment for covariables was used to compare the mean PFM strength and identify its associated factors. Results. The mean PFM strength of women undergoing vaginal delivery and cesarean section was 25.6 +/- 14.5 cmH(2)O and 39.6 +/- 22.0 cmH(2)O (p < 0.01, adjusted for covariables), respectively. A correlation was observed between measurements of PFM strength obtained by vaginal digital palpation and use of a perineometer (tau = 0.82; p < 0.01). The non-white race/ethnicity was negatively associated with PFM strength (coefficient: -10.2424; p = 0.02). Conclusions. A lower PFM strength was observed in women with a history of vaginal delivery compared to those undergoing cesarean section. Non-white race/ethnicity negatively affected PFM strength. Our data suggest that vaginal digital palpation may be used in clinical practice because of its expressive correlation with use of a perineometer.

Comparison of Cross-Sectional Hardness and Transverse Microradiography of Artificial Carious Enamel Lesions Induced by Different Demineralising Solutions and Gels

The aims of this study were: (1) to correlate surface (SH) and cross-sectional hardness (CSH) with microradiographic parameters of artificial enamel lesions; (2) to compare lesions prepared by different protocols. Fifty bovine enamel specimens were allocated by stratified randomisation according to their initial SH values to five groups and lesions produced by different methods: MC gel (methylcellulose gel/lactic acid, pH 4.6, 14 days); PA gel (polyacrylic acid/lactic acid/hydroxyapatite, pH 4.8, 16 h); MHDP (undersaturated lactate buffer/methyl diphosphonate, pH 5.0, 6 days); buffer (undersaturated acetate buffer/fluoride, pH 5.0, 16 h), and pH cycling (7 days). SH of the lesions (SH(1)) was measured. The specimens were longitudinally sectioned and transverse microradiography (TMR) and CSH measured at 10- to 220-mu m depth from the surface. Overall, there was a medium correlation but non-linear and variable relationship between mineral content and root CSH. root SH(1) was weakly to moderately correlated with surface layer properties, weakly correlated with lesion depth but uncorrelated with integrated mineral loss. MHDP lesions showed the highest subsurface mineral loss, followed by pH cycling, buffer, PA gel and MC gel lesions. The conclusions were: (1) CSH, as an alternative to TMR, does not estimate mineral content very accurately, but gives information about mechanical properties of lesions; (2) SH should not be used to analyse lesions; (3) artificial caries lesions produced by the protocols differ, especially considering the method of analysis. Copyright (C) 2009 S. Karger AG, Basel