RESUMEN

List-type questions, which can have a varying number of answers, are more common in the health domain where people seek for health-related information from a passage or passages. An example of this type of question answering task is to find COVID-19 symptoms from a Twitter post. However, due to the lack of annotated instances for supervised learning, automatic identification of COVID-19 symptoms from Twitter posts is challenging. We investigated detection of symptom mentions in Twitter posts using GPT-3, a pre-trained large language model, along with few-shot learning. Our results of 5-shot and 10-shot learning on a corpus of 655 annotated tweets demonstrate that few-shot learning with pre-trained large language model is a promising approach to answering list-type questions with a minimal amount of effort of annotation.

Asunto(s)

RESUMEN

Retrieving health information is a task of search for health-related information from a variety of sources. Gathering self-reported health information may help enrich the knowledge body of the disease and its symptoms. We investigated retrieving symptom mentions in COVID-19-related Twitter posts with a pretrained large language model (GPT-3) without providing any examples (zero-shot learning). We introduced a new performance measure of total match (TM) to include exact, partial and semantic matches. Our results show that the zero-shot approach is a powerful method without the need to annotate any data, and it can assist in generating instances for few-shot learning which may achieve better performance.

Asunto(s)

RESUMEN

Recently, an active area of research in pharmacovigilance is to use social media such as Twitter as an alternative data source to gather patient-generated information pertaining to medication use. Most of thr published work focuses on identifying mentions of adverse effects in social media data but rarely investigating the relationship between a mentioned medication and any mentioned effect expressions. In this study, we treated this relation extraction task as a classification problem, and represented the Twitter text with neural embedding which was fed to a recurrent neural network classifier. The classification performance of our method was investigated in comparison with 4 baseline word embedding methods on a corpus of 9516 annotated tweets.

Asunto(s)

RESUMEN

Since the beginning of the COVID-19 pandemic, patients shared their personal experiences of the viral infection on social media. Gathering their symptomatic experiences reported on Twitter may help better understand the infectious disease and supplement our knowledge of the disease gathered by healthcare workers. In this study, we identified personal experience tweets related to COVID-19 infection using a pre-trained and fine-tuned language model, and annotated the machine-identified tweets in order to extract the information of infection status, symptom concepts, and the days the symptomatic experience occurred. Our result shows that the top 10 most common symptoms mentioned in the collected Twitter data are in line with those published by WHO and CDC. The symptoms along with the day information appear to provide additional insight on how the infection progresses in infected individuals.

Asunto(s)

RESUMEN

HUMIRA, a biologic therapy, has been approved to treat autoimmune diseases and been marketed in many countries worldwide. Much like other medications, it demonstrates many effects on the human body. It is important to understand its effects from the information generated by its users, and social media is one of the venues its users share their experience with the medication. To understand what HUMIRA effects were reported on Twitter, we utilized a relational similarity-based approach to infer HUMIRA effects based upon known medication-effect relations of other medications. With a corpus of 3.6 million preprocessed, "clean" tweets, a total of 55 effects were identified, and among them, 46 were previously observed, and nine were potentially unreported after verification with six reliable sources. The results not only indicate that many HUMIRA effects shared by the Twitter users are consistent with those previously reported, but also demonstrate the power and utility of our method, making it applicable to studying effects of other medications shared by Twitter users.

Asunto(s)

RESUMEN

Continuous monitoring the safe use of medication is an important task in pharmacovigilance. The first-hand experiences of medication effects come from the consumers of the pharmaceuticals. Social media have been considered as a possible alternative data source for gathering consumer-generated information of their experience with medications. Identifying personal experience in social media data is a challenging task in natural language processing. In this study, we investigated a method of predicating personal experience tweets using Google's Bidirectional Encoder Representations from Transformers (BERT) and neural networks, in which BERT models contextually represented the tweet text. Both pre-trained BERT models and our BERT model trained with 3.2 million unlabeled tweets were examined. Our results show that our trained BERT model performs better than Google's pre-trained models (p <; 0.01). This suggests that domain-specific data may contribute to the BERT model yielding better classification performance in predicting personal experience tweets of medication use.

Asunto(s)

RESUMEN

Prevalence of social media has driven a growing number of health related applications with the information shared by online users. It is well known that a gap exists between healthcare professionals and laypeople in expressing the same health concepts. Filling this gap is particularly important for health related applications using social media data. A data-driven, attributional similarity-based method was developed to identify Twitter terms related to side effect concepts. For the 10 most common side effect (symptom) concepts, our method was able to identify a total of 333 Twitter terms, among which only 90 are mapped to those in the consumer health vocabulary (CHV). The identified Twitter terms are specific to Twitter data, indicating a need to expand the existing CHV, and many of them seem to have less ambiguity in word senses than those in CHV.

Asunto(s)

RESUMEN

BACKGROUND: As Twitter has become an active data source for health surveillance research, it is important that efficient and effective methods are developed to identify tweets related to personal health experience. Conventional classification algorithms rely on features engineered by human domain experts, and engineering such features is a challenging task and requires much human intelligence. The resultant features may not be optimal for the classification problem, and can make it challenging for conventional classifiers to correctly predict personal experience tweets (PETs) due to the various ways to express and/or describe personal experience in tweets. In this study, we developed a method that combines word embedding and long short-term memory (LSTM) model without the need to engineer any specific features. Through word embedding, tweet texts were represented as dense vectors which in turn were fed to the LSTM neural network as sequences. RESULTS: Statistical analyses of the results of 10-fold cross-validations of our method and conventional methods indicate that there exist significant differences (p < 0.01) in performance measures of accuracy, precision, recall, F1-score, and ROC/AUC, demonstrating that our approach outperforms the conventional methods in identifying PETs. CONCLUSION: We presented an efficient and effective method of identifying health-related personal experience tweets by combining word embedding and an LSTM neural network. It is conceivable that our method can help accelerate and scale up analyzing textual data of social media for health surveillance purposes, because of no need for the laborious and costly process of engineering features.

Asunto(s)

RESUMEN

Twitter, as a microblogging social media platform, has seen increasing applications of its data for pharmacovigilance which is to monitor and promote safe uses of pharmaceutical products. Medication names are typically used as keywords to query social media data. It is known that medication names are misspelled on social media, and finding the misspellings is challenging because there exists no a priori knowledge as to how people would misspell a medication name. We developed a data-driven, relational similarity-based approach to discover misspellings of medication names. Our approach is based upon the assumption of the identical (or similar) association of a medicine with its effects whether the medication is correctly spelled or misspelled. With distributed representations of the words in tweets posted in recent 24 months, we were able to discover a total of 54 misspellings of 6 medicines whose indications containing headache. Our search results also show that Twitter posts with misspellings of codeine and ibuprofen can be more than 10% of all the tweets associated with each of the medicines. Compared with the phonetics-based approach, our method discovered more actual misspellings used on Twitter.

Asunto(s)

RESUMEN

Health surveillance is an important task to track the happenings related to human health, and one of its areas is pharmacovigilance. Pharmacovigilance tracks and monitors safe use of pharmaceutical products. Pharmacovigilance involves tracking side effects that may be caused by medicines and other health related drugs. Medical professionals have a difficult time collecting this information. It is anticipated that social media could help to collect this data and track side effects. Twitter data can be used for this task given that users post their personal health related experiences on-line. One problem with Twitter data, however, is that it contains a lot of noise. Therefore, an approach is needed to remove the noise. In this paper, several machine learning algorithms including deep neural nets are used to build classifiers that can help to detect these Personal Experience Tweets (PETs). Finally, we propose a method called the Deep Gramulator that improves results. Results of the analysis are presented and discussed.

RESUMEN

BACKGROUND: Audit trails have been used widely to ensure quality of study data and have been implemented in computerized clinical trials data systems. Increasingly, there is a need to audit access to study participant identifiable information to provide assurance that study participant privacy is protected and confidentiality is maintained. In the United States, several federal regulations specify how the audit trail function should be implemented. PURPOSE: To describe the development and implementation of a comprehensive audit trail system that meets the regulatory requirements of assuring data quality and integrity and protecting participant privacy and that is also easy to implement and maintain. METHODS: The audit trail system was designed and developed after we examined regulatory requirements, data access methods, prevailing application architecture, and good security practices. RESULTS: Our comprehensive audit trail system was developed and implemented at the database level using a commercially available database management software product. It captures both data access and data changes with the correct user identifier. Documentation of access is initiated automatically in response to either data retrieval or data change at the database level. LIMITATIONS: Currently, our system has been implemented only on one commercial database management system. Although our audit trail algorithm does not allow for logging aggregate operations, aggregation does not reveal sensitive private participant information. Careful consideration must be given to data items selected for monitoring because selection of all data items using our system can dramatically increase the requirements for computer disk space. Evaluating the criticality and sensitivity of individual data items selected can control the storage requirements for clinical trial audit trail records. CONCLUSIONS: Our audit trail system is capable of logging data access and data change operations to satisfy regulatory requirements. Our approach is applicable to virtually any data that can be stored in a relational database.

Asunto(s)

RESUMEN

Comparative study of biological pathway structures and composition can aid us in elucidating the functions of newly discovered pathways, understanding evolutionary traits, and determining missing pathway elements. A method has been developed to perform pair-wise comparison and similarity search of biological pathways. The comparison determines the differences of each pair of pathways represented in the XML format. The similarity search uses a scoring mechanism to rank the similarities of the pathway in question against those in the pathway repository. To achieve a reasonably good performance, the method is being implemented using the Condor high performance computing environment.

Asunto(s)

RESUMEN

The growing accumulation of biological pathway datasets is making it possible to study pathways at genomic scale by exploration and visualization, facilitating the shift of biomedical research to systems-level analyses. Presented in this paper is our ongoing work on developing a system for exploring and visualizing biological pathways for large scale pathway analyses. A pathway repository with a standard data format was developed for storing various types of pathways, mechanisms for navigating among different pathway datasets and to external data sources were also designed, and a rendering engine is being developed to allow user for visualization and exploration.

Asunto(s)

RESUMEN

The study of biological systems has accumulated a significant amount of biological pathway data, which is evident through the continued growth in both the number of databases and amount of data available. The development of BioPAX standard leads to the increased availability of biological pathway datasets through the use of a special XML format, but the lack of standard storage mechanism makes the querying and aggregation of BioPAX compliant data challenging. To address this shortcoming, we have developed a storage mechanism leveraging the existing XML technologies: the XML database and XQuery. The goal of our project is to provide a generic and centralized store with efficient queries for the needs of biomedical research. A SOAP-based Web service and direct HTTP request methods have also developed to facilitate public consumption of the datasets online.

Asunto(s)

RESUMEN

The massive accumulation of biological data in the past decades has generated a significant amount of biological knowledge which is represented in one way as biological pathways. The existence of over 150 pathway databases reflects the diversity of the biological data and heterogeneity of data models, storage formats and access methods. To address an intriguing biological question, it is not uncommon for a biologist to query more one pathway database to acquire a more complete picture of current understanding of biology. To facility life scientists in searching biological pathway data, we designed a biological pathway aggregator which aggregates various pathway datasets via the BioPAX ontology, a community-developed ontology based upon the concept of Semantic Web for integrating and exchanging biological pathway data. Our aggregator is composed of modules that retrieve the data from various sources, transform the raw data to BioPAX format, persist the converted data in the persistent data store, and enable queries by other applications.

RESUMEN

Conducting clinical trials involves processing a large amount of data generated at different stages of the studies. Such data come from a variety of sources and numerous decisions are made based upon the data to ensure the quality and integrity of the study. The multiplicity of data source poses challenges in making decisions in a timely manner. We designed a solution, called Clinical Trial Console, using Web services, an open standard for programmatic interfaces over the internet based upon XML and other internet protocols, to provide an integrated view of clinical trial data that facilitates prompt decision making.

RESUMEN

Thanks to the advent of information technologies, the emergence of commercial products for electronic clinical trials has improved many aspects of conducting clinical trials. While there are many new options available to facilitate the collection of patient data, little advancements have been made in the way in which information generated from the coordination of a clinical trial is managed. The coordination of a clinical trial has proven to have a significant impact on the quality and economy of clinical trials. The Vanderbilt Coordinating Center has designed and implemented a communication log system (CommLog) to streamline the coordination of clinical trials in order to improve the quality and economy of clinical trials. The CommLog has been operational for several industry-sponsored phase II/III clinical trails and has provided a knowledge base for the studies and repository for useful study information.

Asunto(s)