Search

Social media has grown to be a crucial information source for pharmacovigilance studies where an increasing number of people post adverse reactions to medical drugs that are previously unreported. Aiming to effectively monitor various aspects of Adverse Drug Reactions (ADRs) from diversely expressed social medical posts, we propose a multi-task neural network framework that learns several tasks associated with ADR monitoring with different levels of supervisions collectively. Besides being able to correctly classify ADR posts and accurately extract ADR mentions from online posts, the proposed framework is also able to further understand reasons for which the drug is being taken, known as 'indication', from the given social media post. A coverage-based attention mechanism is adopted in our framework to help the model properly identify 'phrasal' ADRs and Indications that are attentive to multiple words in a post. Our framework is applicable in situations where limited parallel data for different pharmacovigilance tasks are available.We evaluate the proposed framework on real-world Twitter datasets, where the proposed model outperforms the state-of-the-art alternatives of each individual task consistently., Accepted in the research track of The Web Conference(WWW) 2018

Network embedding aims at projecting the network data into a low-dimensional feature space, where the nodes are represented as a unique feature vector and network structure can be effectively preserved. In recent years, more and more online application service sites can be represented as massive and complex networks, which are extremely challenging for traditional machine learning algorithms to deal with. Effective embedding of the complex network data into low-dimension feature representation can both save data storage space and enable traditional machine learning algorithms applicable to handle the network data. Network embedding performance will degrade greatly if the networks are of a sparse structure, like the emerging networks with few connections. In this paper, we propose to learn the embedding representation for a target emerging network based on the broad learning setting, where the emerging network is aligned with other external mature networks at the same time. To solve the problem, a new embedding framework, namely "Deep alIgned autoencoder based eMbEdding" (DIME), is introduced in this paper. DIME handles the diverse link and attribute in a unified analytic based on broad learning, and introduces the multiple aligned attributed heterogeneous social network concept to model the network structure. A set of meta paths are introduced in the paper, which define various kinds of connections among users via the heterogeneous link and attribute information. The closeness among users in the networks are defined as the meta proximity scores, which will be fed into DIME to learn the embedding vectors of users in the emerging network. Extensive experiments have been done on real-world aligned social networks, which have demonstrated the effectiveness of DIME in learning the emerging network embedding vectors., 10 pages, 9 figures, 4 tables. Full paper is accepted by ICDM 2017, In: Proceedings of the 2017 IEEE International Conference on Data MiningS

This study aimed to investigate the clinical efficacy of vacuum-assisted closure (VAC) combined with grafting of artificial dermis and autologous epidermis in the repair of refractory wounds. Patients with refractory wounds underwent debridement. Then the VAC device was used to culture wound granulation tissue. After the wound granulation tissue began to grow, artificial dermis was grafted on the wounds with VAC treatment. Then autologous epidermis was grafted on the artificial dermis to repair the wounds after survival of the artificial epidermis. The study mainly observed length of the hospital stay, survival of the artificial dermis, time required for culture of the granulation tissue using VAC before grafting of the artificial dermis, survival time of the artificial dermis, survival conditions of the autologous epidermis, influence on functions of a healed wound at a functional part, healing conditions of donor sites, and recurrence conditions of the wounds. Healing was successful for 22 patients (95.7%), but treatment failed for 1 child. The 22 patients were followed up for 6 to 24 months. According to follow-up findings, the skin grafts had good color and a soft texture. They were wear resistant and posed no influence on function. The appearance of the final results was the same as that of the full-thickness skin graft. Mild or no pigmentation and no scar formation occurred at the donor sites, and the wounds did not recur. Vacuum-assisted closure combined with grafting of artificial dermis and autologous epidermis is an effective means for repairing refractory wounds and is worth clinical popularizing and application. This journal requires that authors assign a level of evidence to each article. For a full description of these Evidence-Based Medicine ratings, please refer to the Table of Contents or the online Instructions to Authors www.springer.com/00266 .

The Internet has revolutionized healthcare by offering medical information ubiquitously to patients via web search. The healthcare status, complex medical information needs of patients are expressed diversely and implicitly in their medical text queries. Aiming to better capture a focused picture of user's medical-related information search and shed insights on their healthcare information access strategies, it is challenging yet rewarding to detect structured user intentions from their diversely expressed medical text queries. We introduce a graph-based formulation to explore structured concept transitions for effective user intent detection in medical queries, where each node represents a medical concept mention and each directed edge indicates a medical concept transition. A deep model based on multi-task learning is introduced to extract structured semantic transitions from user queries, where the model extracts word-level medical concept mentions as well as sentence-level concept transitions collectively. A customized graph-based mutual transfer loss function is designed to impose explicit constraints and further exploit the contribution of mentioning a medical concept word to the implication of a semantic transition. We observe an 8% relative improvement in AUC and 23% relative reduction in coverage error by comparing the proposed model with the best baseline model for the concept transition inference task on real-world medical text queries., Comment: 10 pages, 2017 IEEE International Conference on Big Data (Big Data 2017)

Given a health-related question (such as "I have a bad stomach ache. What should I do?"), a medical self-diagnosis Android inquires further information from the user, diagnoses the disease, and ultimately recommend best solutions. One practical challenge to build such an Android is to ask correct questions and obtain most relevant information, in order to correctly pinpoint the most likely causes of health conditions. In this paper, we tackle this challenge, named "relevant symptom question generation": Given a limited set of patient described symptoms in the initial question (e.g., "stomach ache"), what are the most critical symptoms to further ask the patient, in order to correctly diagnose their potential problems? We propose an augmented long short-term memory (LSTM) framework, where the network architecture can naturally incorporate the inputs from embedding vectors of patient described symptoms and an initial disease hypothesis given by a predictive model. Then the proposed framework generates the most important symptom questions. The generation process essentially models the conditional probability to observe a new and undisclosed symptom, given a set of symptoms from a patient as well as an initial disease hypothesis. Experimental results show that the proposed model obtains improvements over alternative methods by over 30% (both precision and mean ordinal distance).

A rapid chemiluminescence (CL) reaction was observed when Ce(IV) was injected into the mixed solution of pyridoxine hydrochloride and sodium sulfite. A new flow-injection CL method for the determination of pyridoxine hydrochloride was established. The detection limit (3σ) was 0.5 mg/L pyridoxine hydrochloride and the relative standard deviation (RSD) was 3.3% (n = 11, γ = 5 mg/L). The CL intensity responded linearly to the concentration of pyridoxine hydrochloride in the ranges 1.0–10.0 mg/L and 10.0–100.0 mg/L with linear correlation coefficients of 0.9994 and 0.9991, respectively. The method had been applied to the determination of pyridoxine hydrochloride in tablets and injections successfully.

The low efficiency of hydrolysis is usually the rate-limiting step in the anaerobic digestion of lignocellulosic substrates. In this investigation, bioaugmentation and pretreatment using the cellulolytic anaerobic bacterium Clostridium thermocellum have been found to be beneficial for the improvement of the hydrolysis of switchgrass and enhancing methane production. By adopting these two methods, the methane production of switchgrass reached 289.309 and 278.34 ml/g volatile solids, with an increase of 24.2% and 19.5%, respectively, compared to the blank group. The outcome of this study clearly indicated that the fermentation efficiency could be improved by the use of microorganisms in the above two methods. The methane yield of bioaugmentation not only was slightly higher than that of biopretreatment but also had a higher efficiency in terms of total time and was more adaptive to the industrial implementation. Moreover, this study proved that Clostridium thermocellum required about one week to acclimatize the anaerobic condition and the optimum adding time was at the beginning of anaerobic digestion.

Two-dimensional code has shown its impressive performance on mobile payment and advertising marketing. To ensure that the two-dimensional code is correctly identified by recognition devices, the traditional two-dimensional code must be attached to the surface of plane. However, if the code attached to the surface of cylinder or prism carrier, the image of the normal two-dimensional code will be distortion because of image projection, which will lead to identification problem. To this end, we propose a novel approach of design and identification of anti-distortion two-dimensional code on prism and cylinder combination surface based on pre-stretching, to address the challenge of identifying a two-dimensional code on prism and cylinder combination carrier. According to the spatial geometric relation of the two-dimensional code carrier, pre-stretching the normal two-dimensional code image via inverse mapping to obtain the pre-deformed two-dimensional code that will be printed and then pasted on the surface of the carrier. Specifically, the image of the pre-stretched two-dimensional code in the projection plane of the acquisition device is just as the normal two-dimensional code, which will be correctly identified. Experimental results show that the two-dimensional code after pre-stretching could be identified by normal recognition devices.

The 21st century is the artificial intelligence age; the new media in this condition quickly appeared people's eyes and become the information dissemination and the exchange important platform. In the 2011, people went into the "WeChat age", such as how to obtain the data in the WeChat public platform for interesting data has become the focus of the current research. This article is mainly aimed at the WeChat of the public platform Crawler search technology. First of all, the paper introduces the importance of web crawler in the age of data explosion, and then in the second part of the paper summarizes the domestic and foreign micro-trust Platform Network Crawler technology research, and then a brief overview of the network crawler related theory, working principle and Python in the field of the development of web crawler advantages, Finally, a Python-based WeChat public platform reptile system was designed to successfully obtain data. The crawled data can be used for market research and data analysis as raw data for machine learning and data mining.

Objective To evaluate the effectiveness of pharmaconutrition-supplemented parenteral nutrition (PN) for severe acute pancreatitis (SAP). Methods A comprehensive search of abstracts was performed in the MEDLINE, OVID, Springer, and Cochrane Library database. Published data of randomized clinical trials (RCTs) comparing the clinically relevant outcomes of pharmaconutrition-supplemented PN versus PN for patients with SAP were analyzed. The analyzed outcome variables included infection, mortality, intensive care unit (ICU) stay, hospital stay, and leukocytes change. Statistical analyses were performed using the Cochrane Collaboration’s RevMan 5.1 software. Results Four RCTs published in 1998 or later were included in this meta-analysis, in which 76 patients with pharmaconutrition-supplemented PN and 77 patients with PN. Pharmaconutrition-supplemented PN showed significantly better results in terms of infection (OR, 0.42; 95% CI, 0.20–0.91; P =0.03) and leukocytes change (before treated: mean different, 0.93; 95% CI, 0.21–1.65; P =0.01; after treated: mean different, -0.77; 95% CI, -1.47– -0.08; P =0.03). No significant difference could be found in mortality (OR, 0.30; 95% CI, 0.07–1.19; P =0.09), ICU stay (mean different, -3.65; 95% CI, -9.39–2.10; P =0.21), and hospital stay (mean different, -1.20; 95% CI, -9.89–7.48; P =0.79). Conclusions The current meta-analysis indicates that pharmaconutrition-supplemented PN only show advantages in infection and leukocytes change. Image: Forest plot for infection., JOP. Journal of the Pancreas, Vol 15, No 4 (2014): July - p. 280-412

This paper describes a query-based composition algorithm that can integrate an ARPA format language model in the unified WFST framework, which avoids the memory and time cost of converting the language models to WFST and optimizing the WFST of language models. The proposed algorithm is applied to on-the-fly one-pass decoder and rescoring decoder. Both modified decoder require less memory during decoding on different scale of language models. What’s more, query-based on-the-fly one-pass decoder nearly has the same decoding speed as standard one and query-based rescoring decoder even use less time to rescore the lattice. Because of these advantages, large-scale language models can be applied by query-based composition algorithm to improve the performance of large vocabulary continuous speech recognition.

Electric torque-loading test system can check the disturbed operating performance of servo system by simulating the disturbance torques during the operation. Using LabVIEW software, the PC software of this loading system has been designed in this article. Based on sine wave, square wave and some other waves, this paper focuses on the generation methods of bidirectional square wave and bidirectional triangle wave. The running performance fits the requirements well and the servo system can be effectively tested with the help of the designed simulation torque-loading system.

A continuous-wave high-power single-frequency laser with dual-wavelength output at 1064 and 532 nm is presented. The dependencies of the output power on the transmission of the output coupler and the phase-matching temperature of the LiB(3)O(5) (LBO) crystal are studied. An output coupler with transmission of 19% is used, and the temperature of LBO is controlled to the optimal phase-matching temperature of 422 K; measured maximal output powers of 33.7 W at 1064 nm and of 1.13 W at 532 nm are obtained with optical-optical conversion efficiency of 45.6%. The laser can be single-frequency operated stably and mode-hop-free, and the measured frequency drift is less than 15 MHz in 1 min. The measured Mx2 and My2 for the 1064 nm laser are 1.06 and 1.09, respectively. The measured Mx2 and My2 for the 532 nm laser are 1.12 and 1.11, respectively.

Distributed representations of medical concepts have been used to support downstream clinical tasks recently. Electronic Health Records (EHR) capture different aspects of patients' hospital encounters and serve as a rich source for augmenting clinical decision making by learning robust medical concept embeddings. However, the same medical concept can be recorded in different modalities (e.g., clinical notes, lab results)-with each capturing salient information unique to that modality-and a holistic representation calls for relevant feature ensemble from all information sources. We hypothesize that representations learned from heterogeneous data types would lead to performance enhancement on various clinical informatics and predictive modeling tasks. To this end, our proposed approach makes use of meta-embeddings, embeddings aggregated from learned embeddings. Firstly, modality-specific embeddings for each medical concept is learned with graph autoencoders. The ensemble of all the embeddings is then modeled as a meta-embedding learning problem to incorporate their correlating and complementary information through a joint reconstruction. Empirical results of our model on both quantitative and qualitative clinical evaluations have shown improvements over state-of-the-art embedding models, thus validating our hypothesis., Comment: 9 pages