What is the diagnostic value of ultrasonography compared to physical evaluation in patients with idiopathic carpal tunnel syndrome?

[EN] OBJECTIVE: Our hypothesis is that sonography performed by the rheumatologist in patients with suspected carpal tunnel syndrome (CTS) has higher diagnostic value compared to physical evaluation. METHODS: Adult patients with suspected idiopathic CTS, defined by sensory symptoms over the distribution of the median nerve with or without positive results with the Phalen and/or the Tinel's maneuvers were included. The diagnosis of CTS was indicated by typical symptoms daily for at least 3 months and a positive nerve conduction study. One rheumatologist unaware of the clinical and electrodiagnostic results performed an ultrasound examination of the median nerve for the area ranging from the inlet to the outlet of the carpal tunnel. Mean cross-sectional area at each level, flattening ratio and bowing of flexor retinaculum were obtained. RESULTS: Sixty-eight patients with 105 affected wrists were examined. Tinel's and Phalen's signs had a closer sensitivity (73% and 67% respectively) and specificity (40% and 30% respectively). The best swelling nerve cut-off by sonography was 9.7 mm2 at the tunnel inlet, with a sensitivity of 86%, a specificity of 48% and accuracy of 77%. A 100% positive predictive value was reached with a cross-sectional area of 13 mm2, involving 33 hands (31% of the whole sample). Maximal cross sectional area and the measurement of flexor retinaculum had an accuracy of 72% and 73% respectively. Combination of physical maneuvers and sonography not yielded more accuracy than cross-sectional area itself. CONCLUSION: In patients with clinical history of idiopathic CTS and positive nerve conduction study, sonography performed by the rheumatologist has higher diagnostic value than physical maneuvers.

Evaluation of LBP and HOG descriptors for clothing attribute description

[EN]In this work an experimental study about the capability of the LBP, HOG descriptors and color for clothing attribute classification is presented. Two different variants of the LBP descriptor are considered, the original LBP and the uniform LBP. Two classifiers, Linear SVM and Random Forest, have been included in the comparison because they have been frequently used in clothing attributes classification. The experiments are carried out with a public available dataset, the clothing attribute dataset, that has 26 attributes in total. The obtained accuracies are over 75% in most cases, reaching 80% for the necktie or sleeve length attributes.