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.

Activation of the shoulder and arm muscles during axial load exercises on a stable base of support and on a medicine ball

The purpose of this study was to compare SEMG activities during axial load exercises on a stable base of support and on a medicine ball (relatively unstable). Twelve healthy male volunteers were tested (x = 23 +/- 7y). Surface EMG was recorded from the biceps brachii, anterior deltoid, clavicular portion of pectoralis major, upper trapezius and serratus anterior using surface differential electrodes. All SEMG data are reported as percentage of RMS mean values obtained in maximal voluntary contractions for each muscle studied. A 3-way within factor repeated measures analysis of variance was performed to compare RMS normalized values. The RMS normalized values of the deltoid were always greater during the exercises performed on a medicine ball in relation to those performed on a stable base of support. The trapezius showed greater mean electric activation amplitude values on the wall-press exercise on a medicine ball, and the pectoralis major on the push-up. The serratus and biceps did not show significant differences of electric activation amplitude in relation to both tested bases of support. Independent of the base of support, none of the studied muscles showed significant differences of electric activation amplitude during the bench-press exercise. The results contribute to the identification of the levels of muscular activation amplitude during exercises that are common in clinical practice of rehabilitation of the shoulder and the differences in terms of type of base of support used. (C) 2006 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.

Clear cell odontogenic carcinoma: case report with immunohistochemical findings adding support to the challenging diagnosis

Clear cell odontogenic carcinoma (CCOC) is a rare odontogenic tumor associated with aggressive clinical behavior, metastasis, and low survival. We report a case of CCOC affecting the mandible of a 39-year-old man. The tumor presented a biphasic pattern composed of clear cell nests intermingled with eosinophilic cells and separated by collagenous stroma. Immunoreactivity to cytokeratin (CK), specifically AE1/AE3 and CK 8, 14, 18, and 19 was found, as well as to epithelial membrane antigen (EMA). The tumor cells were negative for S100 protein, CK 13, vimentin, smooth muscle actin, laminin and type IV collagen. Low labeling indices for the proliferation markers Ki-67 and proliferating cell nuclear antigen and to p53 protein might predict a favorable prognosis for the lesion. A surgical resection was performed, followed by adjuvant radiotherapy. A 2-year follow-up has shown no signs of recurrence. The significance of histochemical and immunohistochemical resources in the correct diagnosis of CCOC is analyzed.

Photoelastic Study of the Support Structures of Distal-Extension Removable Partial Dentures

Purpose: The double system of support, in which the distal-extension removable partial denture adapts, causes inadequate stress around abutment teeth, increasing the possibility of unequal bone resorption. Several ways to reduce or more adequately distribute the stress between abutment teeth and residual ridges have been reported; however, there are no definitive answers to the problem. The purpose of this study was to analyze, by means of photoelasticity, the most favorable stress distribution using three retainers: T bar, rest, proximal plate, I bar (RPI), and circumferential with mesialized rest. Materials and Methods: Three photoelastic models were made simulating a Kennedy Class II inferior arch. Fifteen dentures with long saddles, five of each design, were adjusted to the photoelastic patterns and submitted first to uniformly distributed load, and then to a load localized on the last artificial tooth. The saddles were then shortened and the tests repeated. The quantitative and qualitative analyses of stress intensity were done manually and by photography, respectively. For intragroup analyses the Wilcoxon test for paired samples was used, while for intergroup analyses Friedman and Wilcoxon tests were used to better identify the differences (p < 0.05). Results: The RPI retainer, followed by the T bar, demonstrated the best distribution of load between teeth and residual ridge. The circumferential retainer caused greater concentration of stress between dental apexes. Stress distribution was influenced by the type of retainer, the length of the saddle, and the manner of load application. Conclusions: The long saddles and the uniformly distributed loads demonstrated better distribution of stress on support structures.