Half a century ago, the term "computer-aided diagnosis" (CAD) was introduced in the scientific literature. Pulmonary imaging, with chest radiography and computed tomography, has always been one of the focus areas in this field. In this study, I describe how machine learning became the dominant technology for tackling CAD in the lungs, generally producing better results than do classical rule-ba...

background: periodontal infections, which serve as a reservoir of inflammatory mediators, may pose a threat to the fetal-placental unit and cause adverse pregnancy outcomes. objective: the aim of this study was assessing the periodontal status of women during puerperium and determining the possible relationship between their periodontal disease and low birth weight delivery. materials and metho...

BACKGROUND Pulmonary thromboembolism is a common cause of death in patients with autopsy-confirmed Parkinsonism. This study investigated the incidence of leg deep vein thrombosis in Parkinson's disease and relationships between deep vein thrombosis and clinical/laboratory findings, including postural abnormalities as assessed by photographic measurements. METHODS This cross-sectional study as...

purpose: to determine factors that affect and predict final visual acuity, retinal attachment, and ocular survival after open-globe injuries. methods: one-hundred sixteen eyes with open globe injuries were evaluated longitudinally from september 2001to february 2004. age, sex, involved eye, best corrected visual acuity (bcva), afferent pupillary defect (apd), size and location of laceration, ca...

lack of sewage collection systems, percolation of surface waters, and seepage of wells have raised the groundwater table in shiraz area in the south of iran. the growing population generates environmental pollution resulting in the degradation of the quality of surface and groundwaters used for agriculture. inorganic and organic pollutants have been traced in shiraz water resources. heavy metal...

Increasing the size of a neural network typically improves accuracy but also increases the memory and compute requirements for training the model. We introduce methodology for training deep neural networks using half-precision floating point numbers, without losing model accuracy or having to modify hyperparameters. This nearly halves memory requirements and, on recent GPUs, speeds up arithmeti...