Tree-Structured Models for Efficient Multi-Cue Scene Labeling.

We propose a novel approach to semantic scene labeling in urban scenarios, which aims to combine excellent recognition performance with highest levels of computational efficiency. To that end, we exploit efficient tree-structured models on two levels: pixels and superpixels. At the pixel level, we propose to unify pixel labeling and the extraction of semantic texton features within a single architecture, so-called encode-and-classify trees. At the superpixel level, we put forward a multi-cue segmentation tree that groups superpixels at multiple granularities. Through learning, the segmentation tree effectively exploits and aggregates a wide range of complementary information present in the data. A tree-structured CRF is then used to jointly infer the labels of all regions across the tree. Finally, we introduce a novel object-centric evaluation method that specifically addresses the urban setting with its strongly varying object scales. Our experiments demonstrate competitive labeling performance compared to the state of the art, while achieving near real-time frame rates of up to 20 fps.

Validation of a classification system for causes of death in critical care: an assessment of inter-rater reliability.

OBJECTIVE: Trials in critical care have previously used unvalidated systems to classify cause of death. We aimed to provide initial validation of a method to classify cause of death in intensive care unit patients. DESIGN, SETTING AND PARTICIPANTS: One hundred case scenarios of patients who died in an ICU were presented online to raters, who were asked to select a proximate and an underlying cause of death for each, using the ICU Deaths Classification and Reason (ICU-DECLARE) system. We evaluated two methods of categorising proximate cause of death (designated Lists A and B) and one method of categorising underlying cause of death. Raters were ICU specialists and research coordinators from Australia, New Zealand and the United Kingdom. MAIN OUTCOME MEASURES: Inter-rater reliability, as measured by the Fleiss multirater kappa, and the median proportion of raters choosing the most likely diagnosis (defined as the most popular classification choice in each case). RESULTS: Across all raters and cases, for proximate cause of death List A, kappa was 0.54 (95% CI, 0.49-0.60), and for proximate cause of death List B, kappa was 0.58 (95% CI, 0.53-0.63). For the underlying cause of death, kappa was 0.48 (95% CI, 0.44-0.53). The median proportion of raters choosing the most likely diagnosis for proximate cause of death, List A, was 77.5% (interquartile range [IQR], 60.0%-93.8%), and the median proportion choosing the most likely diagnosis for proximate cause of death, List B, was 82.5% (IQR, 60.0%-92.5%). The median proportion choosing the most likely diagnosis for underlying cause was 65.0% (IQR, 50.0%-81.3%). Kappa and median agreement were similar between countries. ICU specialists showed higher kappa and median agreement than research coordinators. CONCLUSIONS: The ICU-DECLARE system allowed ICU doctors to classify the proximate cause of death of patients who died in the ICU with substantial reliability.

Parental calculation of pediatric paracetamol dose: a randomized trial comparing the Parental Analgesia Slide with product information leaflets.

OBJECTIVES: To compare the ability of parents to calculate and demonstrate the correct paracetamol (acetaminophen) dose, interval, and frequency for their child when using either product information leaflets or the Parental Analgesia Slide. BACKGROUND: Prescribing information provided with over-the-counter medication may be a source of confusion for parents delivering analgesics to children at home. Accurate administration is essential to ensure safe and effective treatment of children's pain or fever. The Parental Analgesia Slide is a new device developed with the objective of improving parental dosing accuracy. METHODS: In this prospective, randomized study, 160 parents accompanying children aged between one and 13 years old were randomly allocated to complete a paracetamol dose calculation and administration questionnaire using one of two sources of prescribing information. Absolute percentage dose error and the number of correct dosage intervals, frequencies, and demonstrated drug volumes were compared. RESULTS: Use of the Parental Analgesia Slide resulted in a reduction in the absolute percentage dose error from a median of 33.3 to 0% (P < 0.001) and an increase in the number of correct dosage intervals and frequencies (59/80 to 70/80, P = 0.046). There was no difference in the number of correctly demonstrated drug volumes (P = 0.082) despite a greater number of parents opting to use an oral syringe rather than a dosing spoon when using the Slide (24/80 to 44/80, P = 0.002). CONCLUSIONS: The Parental Analgesia Slide resulted in improved parental ability to calculate paracetamol dose, interval, and frequency while preserving their ability to demonstrate an accurate drug volume.

Dopplersonographic measurement of global cerebral circulation time using echo contrast-enhanced ultrasound in normal individuals and patients with arteriovenous malformations.

Echo-contrast "bolus tracking" by ultrasound (US) is an exciting new tool to study cerebral haemodynamics. In the present study, a global cerebral circulation time (CCT) was measured by extracranial Doppler as the time difference of contrast bolus arrival between the internal carotid artery and internal jugular vein. A total of 64 healthy volunteers and 9 patients with an angiographically diagnosed arteriovenous malformation (AVM) were studied. CCT in volunteers and patients was calculated as the time interval between the points of 10% rise (CCT(1)) and 90% rise (CCT(3)) of the total intensity increase and between the turning points (CCT(2)) of the resulting time-intensity curves. In the volunteer group, CCT(1) was 5.4 +/- 1.8 s, CCT(2) was 7 +/- 1.3 s and CCT(3) 7.5 +/- 1.8 s. CCT results in the AVM group were 2.8 +/- 2.5 s, 3.0 +/- 1.3 s and 4.5 +/- 2.1 s, respectively, and differed significantly from the controls. For the first time, we could confirm a significant shortening of CCT in patients with cerebral AVM by US. The presented test might become a new, additional tool for AVM evaluation and follow-up of treatment in these patients.