Workflow evaluation of environmental contamination with hazardous drugs during compounding and administration in an UK hospital.

Introduction: Exposure of healthcare workers to hazardous drugs may result in adverse health effects underscoring the importance of validating working procedures and safety precautions to minimise the risk. The objective was to monitor environmental contamination caused by the hazardous drug workflow: from drug vials, compounding process, to patient administration.

Methods: Surface wipe samples were collected from potentially contaminated surfaces in the compounding department and in the administration department.

Hierarchical Training of Deep Neural Networks Using Early Exiting.

Deep neural networks (DNNs) provide state-of-the-art accuracy for vision tasks, but they require significant resources for training. Thus, they are trained on cloud servers far from the edge devices that acquire the data. This issue increases communication cost, runtime, and privacy concerns.

Learning Generative Models Using Denoising Density Estimators.

Learning probabilistic models that can estimate the density of a given set of samples, and generate samples from that density, is one of the fundamental challenges in unsupervised machine learning. We introduce a new generative model based on denoising density estimators (DDEs), which are scalar functions parametrized by neural networks, that are efficiently trained to represent kernel density estimators of the data. Leveraging DDEs, our main contribution is a novel technique to obtain generative models by minimizing the Kullback-Leibler (KL)-divergence directly.

No-biopsy strategy for coeliac disease is applicable in adult patients: a 'real-world' Scottish experience.

Objective: Emergency interim guidance from the British Society for Gastroenterology (BSG) states that a no-biopsy strategy is possible to diagnose coeliac disease (CD) in adults with elevated transglutaminase IgA antibody (TGA-IgA) levels. We aimed to determine if the suggested TGA-IgA ≥10× ULN is safe and robust in making the diagnosis in adult patients in Scotland. We also aimed to establish if any important co-diagnoses would be missed if no biopsy was performed.