BRAX, Brazilian labeled chest x-ray dataset.

Chest radiographs allow for the meticulous examination of a patient's chest but demands specialized training for proper interpretation. Automated analysis of medical imaging has become increasingly accessible with the advent of machine learning (ML) algorithms. Large labeled datasets are key elements for training and validation of these ML solutions. In this paper we describe the Brazilian labeled chest x-ray dataset, BRAX: an automatically labeled dataset designed to assist researchers in the validation of ML models. The dataset contains 24,959 chest radiography studies from patients presenting to a large general Brazilian hospital. A total of 40,967 images are available in the BRAX dataset. All images have been verified by trained radiologists and de-identified to protect patient privacy. Fourteen labels were derived from free-text radiology reports written in Brazilian Portuguese using Natural Language Processing.

The efficacy and safety of zolpidem and zopiclone to treat insomnia in Alzheimer's disease: a randomized, triple-blind, placebo-controlled trial.

No prior studies have evaluated the efficacy and safety of zolpidem and zopiclone to treat insomnia of demented patients. This randomized, triple-blind, placebo-controlled clinical trial used these drugs to treat patients with probable, late onset Alzheimer's dementia (AD) (DSM V and NINCDS-ADRDA criteria) exhibiting insomnia (DSM V criteria and nocturnal NPI scores ≥ 2). Actigraphic records were performed for 7 days at baseline and for 14 days during the treatment period in 62 patients aged 80.5 years in average and randomized at a 1:1:1 ratio for administration of zolpidem 10 mg/day, zopiclone 7.5 mg/day or placebo. Primary endpoint was the main nocturnal sleep duration (MNSD), whereas secondary outcomes were the proportion of the night time slept, awake time after sleep onset (WASO), nocturnal awakenings, total daytime sleep time and daytime naps. Cognitive and functional domains were tested before and after drug/placebo use. Three participants under zopiclone use had intervention interrupted due to intense daytime sedation and worsened agitation with wandering. Zopiclone produced an 81 min increase in MNSD (95% confidence interval (CI): -0.8, 163.2), a 26 min reduction in WASO (95% CI: -56.2, 4.8) and a 2-episode decrease in awakening per night (95% CI: -4.0, 0.4) in average compared to placebo. Zolpidem yielded no significant difference in MNSD despite a significant 22 min reduction in WASO (95% CI: -52.5, 8.3) and a reduction of 1 awakening each night (95% CI: -3.4, 1.2) in relation to placebo. There was a 1-point reduction in mean performance in the symbols search test among zolpidem users (95% CI: -4.1, 1.5) and an almost eight-point reduction in average scores in the digit-symbol coding test among zopiclone users (95% CI: -21.7, 6.2). In summary, short-term use of zolpidem or zopiclone by older insomniacs with AD appears to be clinically helpful, even though safety and tolerance remain issues to be personalized in healthcare settings and further investigated in subsequent trials. This trial was registered in ClinicalTrials.gov Identifier: NCT03075241.

Multi-Directional Ankle Impedance During Standing Postures.

In this study, we estimated the multi-directional ankle mechanical impedance in two degrees-of-freedom (DOF) during standing, and determined how the stiffness, damping, and inertia vary with ankle angle and ankle torque at different postures. Fifteen subjects stood on a vibrating instrumented platform in four stationary postures, while subjected to pulse train perturbations in both the sagittal and frontal planes of motion. The four stationary postures were selected to resemble stages within the stance phase of the gait cycle: including post-heel-strike during the loading response, mid-stance, post-mid-stance, and just before the heel rises from the ground in terminal-stance phase. In general, the ankle stiffness and damping increased in all directions as the foot COP moved forward, and more torque is generated in plantarflexion. Interestingly, the multi-directional ankle impedance during standing showed a similar shape and major tilt axes to the results of non-loaded scenarios. However, there were notable differences in the impedance amplitude when the ankle was not under bodyweight loading. Last, the stiffness during standing had similar amplitudes ranges to the time-varying ankle stiffness during the stance phase of dynamic walking estimated in previous studies. These results have implications on the design of new, less physically intense, biomechanics experiments aimed at people with neuromuscular disorders or other physical impairments who cannot complete a standard gait test.

Time-varying impedance of the human ankle in the sagittal and frontal planes during straight walk and turning steps.

This paper describes the methods and experiment protocols for estimation of the human ankle impedance during turning and straight line walking. The ankle impedance of two human subjects during the stance phase of walking in both dorsiflexion plantarflexion (DP) and inversion eversion (IE) were estimated. The impedance was estimated about 8 axes of rotations of the human ankle combining different amounts of DP and IE rotations, and differentiating among positive and negative rotations at 5 instants of the stance length (SL). Specifically, the impedance was estimated at 10%, 30%, 50%, 70% and 90% of the SL. The ankle impedance showed great variability across time, and across the axes of rotation, with consistent larger stiffness and damping in DP than IE. When comparing straight walking and turning, the main differences were in damping at 50%, 70%, and 90% of the SL with an increase in damping at all axes of rotation during turning.

Personalized Modeling for Prediction with Decision-Path Models.

Deriving predictive models in medicine typically relies on a population approach where a single model is developed from a dataset of individuals. In this paper we describe and evaluate a personalized approach in which we construct a new type of decision tree model called decision-path model that takes advantage of the particular features of a given person of interest. We introduce three personalized methods that derive personalized decision-path models. We compared the performance of these methods to that of Classification And Regression Tree (CART) that is a population decision tree to predict seven different outcomes in five medical datasets. Two of the three personalized methods performed statistically significantly better on area under the ROC curve (AUC) and Brier skill score compared to CART. The personalized approach of learning decision path models is a new approach for predictive modeling that can perform better than a population approach.

Temperature-induced coexistence of a conducting bilayer and the bulk-terminated surface of the topological insulator Bi2Te3.

Topological insulators such as Bi2Se3 and Bi2Te3 have extremely promising transport properties, due to their unique electronic behavior: they are insulators in the bulk and conducting at the surface. Recently, the coexistence of two types of surface conducting channels has been observed for Bi2Se3, one being Dirac electrons from the topological state and the other electrons from a conventional two-dimensional gas. As an explanation for this effect, a possible structural modification of the surface of these materials has been hypothesized. Using scanning tunneling microscopy we have directly observed the coexistence of a conducting bilayer and the bare surface of bulk-terminated Bi2Te3. X-ray crystal truncation rod scattering was used to directly show the stabilization of this epitaxial bilayer which is primarily composed of bismuth. Using this information, we have performed density functional theory calculations to determine the electronic properties of the possible surface terminations. They can be used to understand recent angular resolved photoemission data which have revealed this dual surface electronic behavior.