Blomfenger0447

BACKGROUND Automated de-identification methods for removing protected health information (PHI) from the source notes of the electronic health record (EHR) rely on building systems to recognize mentions of PHI in text, but they remain inadequate at ensuring perfect PHI removal. As an alternative to relying on de-identification systems, we propose the following solutions (1) Mapping the corpus of documents to standardized medical vocabulary (concept unique identifier [CUI] codes mapped from the Unified Medical Language System) thus eliminating PHI as inputs to a machine learning model; and (2) training character-based machine learning models that obviate the need for a dictionary containing input words/n-grams. We aim to test the performance of models with and without PHI in a use-case for an opioid misuse classifier. METHODS An observational cohort sampled from adult hospital inpatient encounters at a health system between 2007 and 2017. A case-control stratified sampling (n = 1000) was performed to build an arate good test characteristics for an opioid misuse computable phenotype that is void of any PHI and performs similarly to models that use PHI. Herein we share a PHI-free, trained opioid misuse classifier for other researchers and health systems to use and benchmark to overcome privacy and security concerns.BACKGROUND Naturally acquired immunity (NAI), which is characterized by protection against overt clinical disease and high parasitaemia, is acquired with age and transmission intensity. The role of NAI on the efficacy of anti-malarial drugs, including artemisinin-based combinations used as the first-line treatment for uncomplicated Plasmodium falciparum, has not been fully demonstrated. This study investigated the role of NAI in response to artemisinin-based combination therapy (ACT), in symptomatic patients living in western Kenya, a high malaria transmission area. METHODS Sera samples from malaria immune participants (n = 105) in a therapeutic efficacy study were assessed for in vitro growth inhibitory activity against the 3D7 strain of P. falciparum using a fluorescent-based growth inhibition assay (GIA). Participants' age and parasite clearance parameters were used in the analysis. Pooled sera from malaria naïve participants (n = 6) with no Plasmodium infection from malaria non-endemic regions of Kenya was used as negative control. RESULTS The key observations of the study were as follows (1) Sera with intact complement displayed higher GIA activity at lower (1%) serum dilutions (p  less then  0.0001); (2) there was significant relationship between GIA activity, parasite clearance rate (p = 0.05) and slope half-life (p = 0.025); and (3) age was a confounding factor when comparing the GIA activity with parasite clearance kinetics. CONCLUSION This study demonstrates for the first time there is synergy of complement, pre-existing immunity, and drug treatment in younger patients with symptomatic malaria in a high-transmission area.BACKGROUND Semantic textual similarity (STS) is a fundamental natural language processing (NLP) task which can be widely used in many NLP applications such as Question Answer (QA), Information Retrieval (IR), etc. It is a typical regression problem, and almost all STS systems either use distributed representation or one-hot representation to model sentence pairs. METHODS In this paper, we proposed a novel framework based on a gated network to fuse distributed representation and one-hot representation of sentence pairs. Some current state-of-the-art distributed representation methods, including Convolutional Neural Network (CNN), Bi-directional Long Short Term Memory networks (Bi-LSTM) and Bidirectional Encoder Representations from Transformers (BERT), were used in our framework, and a system based on this framework was developed for a shared task regarding clinical STS organized by BioCreative/OHNLP in 2018. RESULTS Compared with the systems only using distributed representation or one-hot representation, our method achieved much higher Pearson correlation. Among all distributed representations, BERT performed best. The highest Person correlation of our system was 0.8541, higher than the best official one of the BioCreative/OHNLP clinical STS shared task in 2018 (0.8328) by 0.0213. CONCLUSIONS Distributed representation and one-hot representation are complementary to each other and can be fused by gated network.BACKGROUND Symptomatic or active hydrocephalus in children is linked to an elevation in intracranial pressure (ICP), which is likely to be multifactorial in origin. The CSF outflow resistance, venous sinus resistance and total cerebral blood flow are likely factors in the ICP elevation. The purpose of this paper is to define the incidence, site and significance of venous sinus stenosis and/or cerebral hyperemia in a cohort of children diagnosed with hydrocephalus at a tertiary referral hospital. METHODS The imaging database was reviewed over a 10 year period and the index MRI of all children between the ages of 4 months and 15 years, who were diagnosed with treatment naive hydrocephalus of any type (excluding secondary to tumor) and had magnetic resonance venography (MRV) and flow quantification were selected. Patients were compared with children undergoing an MRI with MRV and flow quantification who were subsequently shown to have no abnormality. The cross-sectional area and circumference of the sinuses were measured at 4 levels. The hydraulic and effective diameters were calculated. An area stenosis of 65% or greater was deemed significant. A total cerebral blood flow greater than two standard deviations above the mean for controls was taken to be abnormal. RESULTS There were a total of 55 children with hydrocephalus compared to 118 age matched control MRV's and 35 control flow quantification studies. A high grade stenosis occurred in 56% of patients but in none of the controls (p  less then  0.0001). The commonest site of narrowing was in the distal sigmoid sinus. Cerebral hyperemia occurred in 13% of patients but did not occur in the controls. TH-Z816 ic50 CONCLUSIONS The elevation in ICP in symptomatic hydrocephalus is multifactorial. Both high grade venous stenosis and cerebral hyperemia are common in childhood hydrocephalus. High grade stenosis was noted to be a risk factor for conservative management failure. Hyperemia was a good prognostic indicator.