Medical image retrieval based on complexity analysis

Texture is one of the most important visual attributes used in image analysis. It is used in many content-based image retrieval systems, where it allows the identification of a larger number of images from distinct origins. This paper presents a novel approach for image analysis and retrieval based on complexity analysis. The approach consists of a texture segmentation step, performed by complexity analysis through BoxCounting fractal dimension, followed by the estimation of complexity of each computed region by multiscale fractal dimension. Experiments have been performed with MRI database in both pattern recognition and image retrieval contexts. Results show the accuracy of the method and also indicate how the performance changes as the texture segmentation process is altered.

End-of-life practices in seven Brazilian pediatric intensive care units

Objective: To evaluate the incidence of life support limitation and medical practices in the last 48 hrs of life of children in seven Brazilian pediatric intensive care units (PICUs). Design. Cross-sectional multicenter retrospective study based on medical chart review. Setting: Seven PICUs belonging to university and tertiary hospitals located in three Brazilian regions: two in Porto Alegre (southern region), two in Sao Paulo (southeastern region), and three in Salvador (northeastern region). Patients. Medical records of all children who died in seven PICUs from January 2003 to December 2004. Deaths in the first 24 hrs of admission to the PICU and brain death were excluded. Interventions: Two pediatric intensive care residents from each PICU were trained to fill out a standard protocol (K = 0.9) to record demographic data and all medical management provided in the last 48 hrs of life (inotropes, sedatives, mechanical ventilation, full resuscitation maneuvers or not). Student`s t-test, analysis of variance, chi-square test, and relative risk were used for comparison of data. Measurements and Main Results. Five hundred and sixty-one deaths were identified; 97 records were excluded (61 because of brain death and 36 due to <24 hrs in the PICU). Thirty-six medical charts could not be found. Cardiopulmonary resuscitation was performed in 242 children (57%) with a significant difference between the southeastern and northeastern regions (p =.0003). Older age (p = .025) and longer PICU stay (p = .001) were associated with do-not-resuscitate orders. In just 52.5% of the patients with life support limitation, the decision was clearly recorded in the medical chart. No ventilatory support was provided in 14 cases. Inotropic drug infusions were maintained or increased in 66% of patients with do-not-resuscitate orders. Conclusions. The incidence of life support limitation has increased among Brazilian PICUs but with significant regional differences. Do-not-resuscitate orders are still the most common practice, with scarce initiatives for withdrawing or withholding life support measures.