No abstract

No abstract

Action recognition in videos is a challenging task due to the complexity of the spatio-temporal patterns to model and the difficulty to acquire and learn on large quantities of video data. Deep learning, although a breakthrough for image classification and showing promise for videos, has still not clearly superseded action recognition methods using hand-crafted features, even when training on massive datasets. In this paper, we introduce hybrid video classification architectures based on carefully designed unsupervised representations of hand-crafted spatio-temporal features classified by supervised deep networks. As we show in our experiments on five popular benchmarks for action recognition, our hybrid model combines the best of both worlds: it is data efficient (trained on 150 to 10000 short clips) and yet improves significantly on the state of the art, including recent deep models trained on millions of manually labelled images and videos.

Introduction Many authors have written about the need to treat patients closer to their beds, in order to observe them more as distinct people. The FAST HUG mnemonic, which consists of a checklist, was suggested as an idea to be employed everyday, by professionals dealing with patients who are critically ill. Minding these questions and motivated by an idea of follow patients' treatment closer, we have put into practice the instrument developed by Jean-Louis Vincent, evaluating the seven most important procedures in critically ill patients, and performed the FAST HUG. This checklist consists of seven items to be evaluated: Feeding, Analgesia, Sedation, Thromboembolic prophylaxis, Head-of-bed elevation, stress Ulcer prevention, and Glucose control. Knowing that the pressure ulcer is one of the challenges faced by ICU nurses, related to patients' need to stay at rest, to be under rigorous control or more complex therapy, it was decided to create the eighth item on the checklist: S, for skin. It stands for skin treatment, with the techniques used in the unit (Braden Scale), monitoring and evaluating closer skin integrity, and allowing nurses to calculate the scoring average of the Braden Scale, and greater incidence of ulcer in interned patients. Objective To expose the shortcomings found during the FAST HUG application, and to show results obtained with the eighth item of the FAST HUG mnemonic: S-Skin. Methods A descriptive study, based on institutional data, was carried out in the adult ICU of a private hospital. It was performed from 2 to 27 June 2008, except on weekends. Three hundred and twenty-three patients were involved. The checklist was carried out during the afternoons by the head nurse, or the assistant nurse of the unit. In order to do this job, a spreadsheet was elaborated to control data, updated every week. This spreadsheet provided graphics for a more objective control of the results obtained. The idea was exposed to the team, during a training program, and so we started the activities. Results and discussion For 20 days of the checklist, 323 patients were evaluated for the eight items. The real shortcomings most frequently found were related to thromboembolic prophylaxis (85%) and glucose control (90%). These shortcomings were immediately evaluated and, depending on this analysis, this item would go on or not, according to the patient's clinical situation. The shortcomings found were tracked just as they were detected, and their cause would be discussed in a multidisciplinary group, and a solution was found. If the item was not observed, it would be 1. Vincent JL: Give your patient a fast hug (at least) once a day.

IntroductionIn coronary artery bypass (CABG) surgery, the common complications are the need for reintubation, prolonged mechanical ventilation (PMV) and death. Thus, a reliable model for the prognostic evaluation of those particular outcomes is a worthwhile pursuit. The existence of such a system would lead to better resource planning, cost reductions and an increased ability to guide preventive strategies. The aim of this study was to compare different methods – logistic regression (LR) and artificial neural networks (ANNs) – in accomplishing this goal.Material and methodsSubjects undergoing CABG (n = 1315) were divided into training (n = 1053) and validation (n = 262) groups. The set of independent variables consisted of age, gender, weight, height, body mass index, diabetes, creatinine level, cardiopulmonary bypass, presence of preserved ventricular function, moderate and severe ventricular dysfunction and total number of grafts. The PMV was also an input for the prediction of death. The ability of ANN to discriminate outcomes was assessed using receiver-operating characteristic (ROC) analysis and the results were compared using a multivariate LR.ResultsThe ROC curve areas for LR and ANN models, respectively, were: for reintubation 0.62 (CI: 0.50–0.75) and 0.65 (CI: 0.53–0.77); for PMV 0.67 (CI: 0.57–0.78) and 0.72 (CI: 0.64–0.81); and for death 0.86 (CI: 0.79–0.93) and 0.85 (CI: 0.80–0.91). No differences were observed between models.ConclusionsThe ANN has similar discriminating power in predicting reintubation, PMV and death outcomes. Thus, both models may be applicable as a predictor for these outcomes in subjects undergoing CABG.

No abstract

No abstract

No abstract