There was no correlation between pain severity and disease severity by sinus CT scan as graded by the Lund-McKay, Harvard, or Kennedy staging system. Facial pain and headache, although frequent complaints of patients with rhinosinusitis, are not useful predictors of sinus disease severity. There appears to be a difference in pain perception between the two North American populations.

Use of the Canadian probability model for DVT in this ED resulted in effective risk stratification, comparable to previously published results.

Objective: To validate the predictive value of the Canadian clinical probability model for acute venous thrombosis, which, to the best of the authors' knowledge, has not been done in emergency department (ED) settings outside of Canada. Methods: Demographic and clinical information, rapid D-dimer testing, and venous ultrasound imaging were obtained among patients presenting with clinically suspected venous thrombosis at a university-affiliated ED. A diagnosis of deep venous thrombosis (DVT) was made based on venous ultrasound test results or objectively documented venous thromboembolism during a 12-week follow-up period. The probability of venous thrombosis was calculated using the Canadian clinical probability model. Results: Among 102 patients, 17 (17%) were diagnosed as having venous thrombosis initially or during the three-month follow-up period. The frequency of venous thrombosis among patients categorized as having high probability was 10 of 17 [59%, 95% confidence interval (95% CI) = 35% to 82%], 6 of 44 (14%, 95% CI = 4% to 24%) with intermediate probability, and 1 of 41 (2%, 95% CI = 0.1% to 11%) with low probability. This compares with respective values of 49%, 14%, and 3%, reported by Canadian researchers in an ED study. Forty-one of 102 (40%) patients had an alternate diagnosis as likely or more likely than venous thrombosis, but only three (7%, 95% CI = 2% to 18%) of these had venous thrombosis. Conclusions: Use of the Canadian probability model for DVT in this ED resulted in effective risk stratification, comparable to previously published results.

No abstract

Within state-of-the-art gesture-based upper-limb myoelectric prosthesis control, gesture recognition commonly relies on the classification of features extracted from electromygraphic (EMG) data gathered from the amputee's residual forearm musculature. Despite best efforts in broadly maximizing gesture recognition accuracy, there does not yet exist a feature-classifier combination accepted as best-practice. In turn, this work hypothesizes that no single feature-classifier combination can consistently maximize accuracy across subjects, positing instead that control schemes should be personalized to the individual. To investigate this hypothesis, the study employed the 40-subject, 49gesture Ninapro DB2 to compare the performance of 7 different historic, more recent and state-of-the-art feature sets, in combination with 5 machine learning classifiers commonly seen within EMG-based pattern recognition literature. The results demonstrate the ability of Linear Discriminant Analysis (LDA) to marginally exceed other more computationally intensive classifiers in terms of mean accuracy, while the feature set which maximized the highest proportion of individuals' accuracies was shown to vary with both classifier choice and gesture count.