Search

Searching, reading, and finding information from the massive medical text collections are challenging. A typical biomedical search engine is not feasible to navigate each article to find critical information or keyphrases. Moreover, few tools provide a visualization of the relevant phrases to the query. However, there is a need to extract the keyphrases from each document for indexing and efficient search. The transformer-based neural networks—BERT has been used for various natural language processing tasks. The built-in self-attention mechanism can capture the associations between words and phrases in a sentence. This research investigates whether the self-attentions can be utilized to extract keyphrases from a document in an unsupervised manner and identify relevancy between phrases to construct a query relevancy phrase graph to visualize the search corpus phrases on their relevancy and importance. The comparison with six baseline methods shows that the self-attention-based unsupervised keyphrase extraction works well on a medical literature dataset. This unsupervised keyphrase extraction model can also be applied to other text data. The query relevancy graph model is applied to the COVID-19 literature dataset and to demonstrate that the attention-based phrase graph can successfully identify the medical phrases relevant to the query terms.

The semantic mismatch between query and document terms—i.e., the semantic gap—is a long-standing problem in Information Retrieval (IR). Two main linguistic features related to the semantic gap that can be exploited to improve retrieval are synonymy and polysemy. Recent works integrate knowledge from curated external resources into the learning process of neural language models to reduce the effect of the semantic gap. However, these knowledge-enhanced language models have been used in IR mostly for re-ranking and not directly for document retrieval. We propose the Semantic-Aware Neural Framework for IR (SAFIR), an unsupervised knowledge-enhanced neural framework explicitly tailored for IR. SAFIR jointly learns word, concept, and document representations from scratch. The learned representations encode both polysemy and synonymy to address the semantic gap. SAFIR can be employed in any domain where external knowledge resources are available. We investigate its application in the medical domain where the semantic gap is prominent and there are many specialized and manually curated knowledge resources. The evaluation on shared test collections for medical literature retrieval shows the effectiveness of SAFIR in terms of retrieving and ranking relevant documents most affected by the semantic gap.

Objectives To identify and assess the evidence showing the efficacy of auricular acupressure alone for myopia in children and adolescents. Methods Randomized controlled trials (RCTs) that were published until March 2019 in Pubmed, Web of Science, OVID, Foreign Medical Literature Retrieval Service, China Knowledge Resource Integrated Database, The Chinese Biological Medicine Database, Wanfang Database, and Chongqing VIP Information were searched. The quality of RCTs was assessed using the Cochrane risk of bias assessment tool. Results Ten RCTs were included to be qualitatively summarized, of which 5 studies qualified for the meta-analysis of the efficacy rate in treating myopia. This review demonstrated that auricular acupressure alone was more effective than eye-drops treatment, eye exercise, and was the just as effective as needle acupuncture. Conclusions Auricular acupressure could slow the progression of myopia in children and adolescents. However, there is a need for further studies with higher methodological quality and sufficient follow-up.

As an important branch of clinical decision support track, medical literature retrieval receives the electronic medical record as an input to obtain the acquired information from a large amount of medical literature. However, the description of clinical record is rather complicated and ambiguous in semantic way, therefore it is necessary for us to perform query reconstruction for a better retrieval performance. In this paper, we take the advantage of query quality predictors to process long clinical notes with redundant content, and predict query intent to reconstruct the original query. Experimental results on the standard Text REtrieval Conference (TREC) CDS track dataset confirm the superior performance of the proposed method.

The purpose of this study is to apply the Jigsaw teaching mode into the medical literature retrieval and discuss the effect of the new teaching one. According to the Jigsaw learning mode, the students are divided into expert and Jigsaw groups. Through segment allocation, discussion in the expert groups and presentation in the Jigsaw groups, students have more initiative and learn more with this method. The result is that Jigsaw is an effective cooperative learning mode and the students' learning enthusiasm and autonomous learning ability can be well promoted in the medical literature retrieval.

Developing skills to search the medical literature has potential benefits on patient care and allow physicians to better orient their efforts when answering daily clinical questions. The objective of this paper is to share useful tools for optimizing medical literature retrieval in MEDLINE using PubMed including MeSH terms, filters and connectors.

The rapid increase of medical literature poses a significant challenge for physicians, who have repeatedly reported to struggle to keep up to date with developments in research. This gap is one of the main challenges in integrating recent advances in clinical research with day-to-day practice. Thus, the need for clinical decision support (CDS) search systems that can retrieve highly relevant medical literature given a clinical note describing a patient has emerged. However, clinical notes are inherently noisy, thus not being fit to be used as queries as-is. In this work, we present a convolutional neural model aimed at improving clinical notes representation, making them suitable for document retrieval. The system is designed to predict, for each clinical note term, its importance in relevant documents. The approach was evaluated on the 2016 TREC CDS dataset, where it achieved a 37% improvement in infNDCG over state-of-the-art query reduction methods and a 27% improvement over the best known method for the task.

Background The constantly growing publication rate of medical research articles puts increasing pressure on medical specialists who need to be aware of the recent developments in their field. The currently used literature retrieval systems allow researchers to find specific papers; however the search task is still repetitive and time-consuming. Aims In this paper we describe a system that retrieves medical publications by automatically generating queries based on data from an electronic patient record. This allows the doctor to focus on medical issues and provide an improved service to the patient, with higher confidence that it is underpinned by current research. Method Our research prototype automatically generates query terms based on the patient record and adds weight factors for each term. Currently the patient’s age is taken into account with a fuzzy logic derived weight, and terms describing blood-related anomalies are derived from recent blood test results. Conditionally selected homonyms are used for query expansion. The query retrieves matching records from a local index of PubMed publications and displays results in descending relevance for the given patient. Recent publications are clearly highlighted for instant recognition by the researcher. Results Nine medical specialists from the Royal Adelaide Hospital evaluated the system and submitted pre-trial and post-trial questionnaires. Throughout the study we received positive feedback as doctors felt the support provided by the prototype was useful, and which they would like to use in their daily routine. Conclusion By supporting the time-consuming task of query formulation and iterative modification as well as by presenting the search results in order of relevance for the specific patient, literature retrieval becomes part of the daily workflow of busy professionals.

Objective This paper presents a study of methods for medical literature retrieval for case queries, in which the goal is to retrieve literature articles similar to a given patient case. In particular, it focuses on analyzing the performance of state-of-the-art general retrieval methods and improving them by the use of medical thesauri and physician feedback. Materials and Methods The Kullback–Leibler divergence retrieval model with Dirichlet smoothing is used as the state-of-the-art general retrieval method. Pseudorelevance feedback and term weighing methods are proposed by leveraging MeSH and UMLS thesauri. Evaluation is performed on a test collection recently created for the ImageCLEF medical case retrieval challenge. Results Experimental results show that a well-tuned state-of-the-art general retrieval model achieves a mean average precision of 0.2754, but the performance can be improved by over 40% to 0.3980, through the proposed methods. Discussion The results over the ImageCLEF test collection, which is currently the best collection available for the task, are encouraging. There are, however, limitations due to small evaluation set size. The analysis shows that further refinement of the methods is necessary before they can be really useful in a clinical setting. Conclusion Medical case-based literature retrieval is a critical search application that presents a number of unique challenges. This analysis shows that the state-of-the-art general retrieval models are reasonably good for the task, but the performance can be significantly improved by developing new task-specific retrieval models that incorporate medical thesauri and physician feedback.

Searching the literature has a direct, beneficial influence on patient care. The amount of medical scientific information has increased to a great extent, while the development of networking technologies has broadened access to online databases. Successful searches depend upon understanding technical librarianship concepts and the skills for mastering searching interfaces. From a problem-oriented approach, concepts like MEDLINE coverage, PubMed resources, Boolean logic, search strategies, and Web sources for full-text articles are introduced along seven online situations: locating a specific publication, answering a complex clinical question, finding information on a general subject, finding publications by a particular author, finding publications in a particular language, finding a specific publication type, and locating the full-text document. Oncologists should face the challenge of performing their own searches. Specific knowledge is mandatory to avoid frustrating, time-consuming work. The objective of this work is to present concepts, strategies, and skills required for medical literature retrieval, easing the incorporation of new and welcomed practices.

The use of electronic information has been increasing in medical science. Medical students need to know how to get the information that they want. Medical librarians are specialists on medical information and must be able to help medical students find what they need. Consequently, medical school librarians should provide user education opportunities, including topics on information literacy in regular classes. Questionnaires were mailed to 80 medical school libraries. 72 libraries (90%) responded. In 45 of these libraries (62%), librarians are presently teaching user education in regular classes and in 8 of them (11%), librarians were teaching before. The major reason why the librarians had started to teach such classes was because of requests from teachers. In 17 of 53 libraries (32%), the librarians felt that teaching regular classes was very hard work because it takes much time to prepare for teaching. Most librarians are teaching medical literature retrieval. The results of the survey seem to show that most medical librarians wrestle with teaching user education in regular classes. Whether librarians continue to teach such topics in regular classes depends on the opinions of teachers and it is an important issue in future.

Visualization of Concepts in Medicine (VCM) is a compositional iconic language that aims to ease information retrieval in Electronic Health Records (EHR), clinical guidelines or other medical documents. Using VCM language in medical applications requires alignment with medical reference terminologies. Alignment from Medical Subject Headings (MeSH) thesaurus and International Classification of Diseases – tenth revision (ICD10) to VCM are presented here. This study aim was to evaluate alignment quality between VCM and other terminologies using different measures of inter-alignment agreement before integration in EHR. For medical literature retrieval purposes and EHR browsing, the MeSH thesaurus and the ICD10, both organized hierarchically, were aligned to VCM language. Some MeSH to VCM alignments were performed automatically but others were performed manually and validated. ICD10 to VCM alignment was entirely manually performed. Inter-alignment agreement was assessed on ICD10 codes and MeSH descriptors, sharing the same Concept Unique Identifiers in the Unified Medical Language System (UMLS). Three metrics were used to compare two VCM icons: binary comparison, crude Dice Similarity Coefficient (DSC crude ), and semantic Dice Similarity Coefficient (DSC semantic ), based on Lin similarity. An analysis of discrepancies was performed. MeSH to VCM alignment resulted in 10,783 relations: 1,830 of which were manually performed and 8,953 were automatically inherited. ICD10 to VCM alignment led to 19,852 relations. UMLS gathered 1,887 alignments between ICD10 and MeSH. Only 1,606 of them were used for this study. Inter-alignment agreement using only validated MeSH to VCM alignment was 74.2% [70.5-78.0]CI95%, DSC crude was 0.93 [0.91-0.94]CI95%, and DSC semantic was 0.96 [0.95-0.96]CI95%. Discrepancy analysis revealed that even if two thirds of errors came from the reviewers, UMLS was nevertheless responsible for one third. This study has shown strong overall inter-alignment agreement between MeSH to VCM and ICD10 to VCM manual alignments. VCM icons have now been integrated into a guideline search engine ( http://www.cismef.org ) and a health terminologies portal ( http://www.hetop.eu ).

The aim of this study was to discuss the research focus on Network Interactive Services in university library in china. We adopt the method of research focus analysis based on the frequency of keyword and g-index, taking database China Academic Journal Database (CAJD) as the data recourse. The results show that development of digital referential consultant service and collaborative model, technology of internet interactive services, knowledge management in the role of the internet interactive service, are the research focus of medical literature retrieval course in the last 7 years.

This paper is a report on library use and user education at the Library of Kochi Medical School. The education programs are “Orientation”, “Library Tour”, “Library Guidance”, “Bibliographic Instruction”, “Information Retrieval Training”, and others. Lectures and practical exercises were started in 1999 in the medical course curriculum. They were first provided for 4th year medical students, then for 3rd year nursing students, and lately for freshmen in both courses. The practical exercises include medical literature retrieval training of MEDLINE, Igaku Chuo Zasshi, and electronic journal systems. It is necessary and important to have programs on information literacy education, particularly in medical libraries, now and in the future.

The business of health delivery is complex. Employing over 850,000 people, and delivering services to 21.3 million residents, the Australian health care system is currently strained to the maximum in dealing with increasing demand for services and an acute shortage of skilled professionals. The National e–Health Strategy drives a nationwide research agenda to provide the infrastructure and tools required to support the planning, management and delivery of health care services. Deriving principles from the disciplines of computer science, mathematics, philosophy and physiology, and consisting of different fields, from machine vision to expert systems, the field of Artificial Intelligence (AI) deals with the creation of "machines that can think". Focused on traits of reasoning, knowledge representation, planning, learning, communication, perception and social intelligence, AI has been widely applied to augment the state of the art in Health Informatics. This special issue reports on the latest developments in the field of AI motivated research in the health domain. The special issue arose from the inaugural Australian Workshop on Artificial Intelligence in Health (AIH 2011). It was held in conjunction with the 24th Australasian Joint Conference On Artificial Intelligence (AI2011), in Perth, Australia, in December 2011. The AIH 2011 workshop was a first of its kind, national initiative that aimed to bring together scholars and practitioners in the field of AI–driven health informatics to present and discuss their research, share their knowledge and experiences, define key research challenges and explore possible collaborations to advance e–Health development nationally and internationally. Therefore, the affiliation of AIH 2011 with AI2011 was both timely and mutually beneficial for the communities involved in these events. AIH 2011 received 16 full paper submissions and each paper was reviewed by three program committee members. Six papers were accepted as full papers and five as short papers accompanied with posters. The workshop brought together researchers from a variety of disciplines across various parts of the country and provided an excellent forum for discussion and exchange of ideas. In addition to presentation of papers, the workshop featured two keynote addresses, a poster session during lunch, and a panel discussion. The first keynote address, “Using Artificial Intelligence to transform the management of Chronic Disease”, was delivered by Professor Michael Georgeff. In addition to presenting ongoing research efforts to apply AI techniques to better manage chronic disease, Professor Georgeff discussed the greatest innovations in healthcare from a medical practitioner’s viewpoint and focussed on how AI could be leveraged to transform healthcare. The second keynote address, “Knowledge Acquisition Issues in Interpreting Laboratory Data”, was presented by Professor Paul Compton and posed a number of questions related to large–scale knowledge acquisition for decision– support systems in medicine, drawing lessons from experience in working with knowledge acquisition tools that support over 300 million diagnostics laboratory reports. The workshop concluded with a panel discussion in which Professor Abdul Sattar and Professor Yogi Kanagasingam joined the keynote speakers to address the topic “AI for eHealth: 2012 and Beyond”. The discussion that ensued was very energetic and actively engaged audience interaction. Topics discussed ranged from the emerging contribution of AI over the past five decades, current issues with the marketing and uptake of AI–based solutions in mainstream healthcare, and the challenges for AI–based research and application in years to come. The panel and audience also acknowledged the key role a workshop like this would play in driving collaborative research efforts between AI and health informatics research communities. All accepted papers were also invited to revise and submit their manuscripts for inclusion in this special issue of the Australasian Medical Journal (AMJ). Seven papers and a letter to the editor have been accepted for publication in the journal. The first paper, by Bevan Koopman, Peter Bruza, Laurianne Sitbon and Michael Lawley, titled “Towards Semantic Search and Inference in Electronic Medical Records”, presents concept–based information retrieval for searching electronic medical records and demonstrates that the approach outperforms keyword–based search, working especially well for queries where the latter performs poorly. This paper was awarded the best paper prize at the workshop. The second paper, by Kinzang Chhogyal, Abhaya Nayak, Rolf Schwitter and Abdul Sattar, titled “A Causal Model for Fluctuating Sugar Levels in Diabetes Patients”, investigates the use of fixed distance based belief revision to improve causal models. A simple scenario for fluctuating blood sugar levels in a diabetes patient is used to demonstrate the efficacy of this approach. The third paper, by Alexander Krumpholz, David Hawking, Richard Jones, Tom Gedeon and Hugh Greville, titled “Automated Medical Literature Retrieval”, describes a system for the retrieval of relevant medical publications using queries generated automatically from data present in an electronic patient record. Integrated into an electronic record system, such a system would proactively support medical practitioners in the delivery of care. The fourth paper, by Abeed Sarker, Diego Molla and Cecile Paris, titled “Extractive Summarisation of Medical Documents”, proposes a query focused approach for automatically summarising medical documents and helping medical practitioners find relevant information. The fifth paper, by Diego Molla and Maria Elena Santiago– Martinez, titled “A Corpus for Evidence Based Summarisation”, presents a corpus of clinical questions and answers designed for training and testing automated text summarisers to support evidence–based medicine. The sixth paper, by Di Xiao, Janardhan Vignarajan, Jane Lock, Shaun Frost, Mei–Ling Tay–Kearney and Yogi Kanagasingam, titled “Retinal Image Registration and Comparison for Clinical Decision Support”, proposes a set of accurate and robust retinal image registration solutions for longitudinal retinal image alignment and comparison The seventh paper, by Amol Wagholikar, Maggie Fung and Colleen Nelson, titled “Improving Self–Care of Patients with Chronic Disease using Online Personal Health Record”, employs a case–based reasoning approach to self care for advanced prostate cancer patients in an online patient health record environment. A letter to the editor, by Anthony Nguyen, Yue Kimi Sun, Laurianne Sitbon and Shlomo Geva, titled “Representation Of Assertions In Clinical Free Text Using SNOMED CT,” explores the use of the SNOMED CT terminology for representing medical concepts and their assertions in clinical free text. The authors demonstrate that that populating assertions as attribute values using SNOMED CT allows over 93% of assertions in their experimental dataset to be represented. We hope that the breadth and diversity of the papers presented at the workshop and published in this special issue will foster further collaboration and AI driven research in health. This workshop would not have been possible without the contributions of numerous fine people. First, we are greatly indebted to Professor Aditya Ghose, Professor Anthony Maeder, Professor Wayne Wobcke, Professor Mehmet Orgun, and Dr Yogesan (Yogi) Kanagasingam for their guidance and support. We would also like to thank the organising committee of the 24th Australasian Joint Conference on AI, the Institute of Integrated and Intelligent Systems, Griffith University for supporting the workshop, and the CSIRO Australian e–Health Research Centre for their support and sponsorship of travel scholarships and the best paper prize. Thanks are also due to Professor Moyez Jiwa and the AMJ for supporting the workshop and inviting accepted papers for inclusion into this special issue. Finally, we are indebted to the authors who responded to the invitation to submit their papers, and the reviewers who generously donated their time and expertise and provided very comprehensive reviews of the submitted papers.

As part of a U.S government effort to combat AIDS, the National Library of Medicine (NLM) is expanding its information services in this area. In response to the Health Omnibus Program Extension Act of 1988, NLM has expanded its AIDS Bibliography, enhanced its DIRLINE file with AIDS-related organizations, and developed three new online MEDLARS (Medical Literature Retrieval and Analysis System) databases: AIDSLINE, AIDSTRIALS, and AIDSDRUGS. NLM's AIDS-related programs and services are discussed.

INTRODUCTION: IATROTEK is a bilingual (Greek, English) database containing the references of about 15.000 articles (abstracts included), which have been published in 75 Greek biomedical journals since 1982. Until recently, the input in this database was performed by manual typing and the retrieval was available to the users only off line. Input automation was started before six months and on-line use of the database was officially inaugurated on May 8th, 1999 (www.mednet.gr/iatrotek). METHODS: The on-line version of IATROTEK is implemented on a Linux-based server running PostgreSQL database server and Apache web server. All the software used is freeware.The user interface is comprised of several HTML forms, which are accessible by any WWW browser, and it is supported by certain CGI (Common Gateway Interface) scripts written in Perl. Searching can be performed providing single terms or terms combined by boolean operators, on the following fields: title, authors, key-words, abstract in both Greek and English. Matching records include the title and reference of the articles as well as links to the Greek and the English abstracts. Links providing further searching on authors' names and key-words are available on the page of abstracts. Concerning the input in the database, it is performed by the use of appropriate filters (format transformers) on the data provided by the publishers. Under construction is the automation of key-word assigning procedure and a stemming algorithm for Greek words. RESULTS: During the first fifteen days of its on-line operation, the database is queried with a mean rate of 48 requests/day (20% of them from abroad). Bearing in mind the small number of Greek doctors having access to the Web (2,500-3,000 out of 35,000), this is a fact signifying their warm interest on the on-line use of the IATROTEK. DISCUSSION: Compared to the availability of the international medical literature through the Web, the literature written in national languages is not equally favored by the Internet explosion. Efforts, like that of IATROTEK on-line, towards the dissemination of locally produced medical information through the Web will overcome the serious obstacles confronted by the national journals. []

It can he projected that in the near future plysicians with personal computers connected to telephone lines will routinely have access directly into imst computer-stored medical databases. 7% will permit eflicient online search and retrieval of the full text of relevant medical literature for practice-linked medical education and clinical decision suppotT Thus, the gap between what is known and what is practiced can he narrowed.

As quantitative bio-medical science develops, the amount of data to be handled will increase. Already the investigator frequently finds that the analysis of the data he has collected is beyond his computational ability and he is forced to turn to automatic data processing. The bio-medical investigator must have access to a variety of both analog and digital computing equipment if progress is to be made in biomedical research. Automatic data-processing techniques and equipment need to be introduced to the areas of public health and patient care. Projects dealing with medical records, medical literature retrieval, automatic chemical analysis, the recording of signs, symptoms, nurses' notes, etc., must all be tied together in regional medical computing systems as soon as possible. Particular attention must be given to this systematic collection and analysis of medical data before regional computing systems can help physicians with diagnosis and treatment.

To the Editor.— The article on full-text medical literature retrieval by Collen and Flagle 1 introduced readers to an important source of medical information, but one question left unanswered in that piece is potentially more important than the ones addressed. Lawyers have been working with full-text document retrieval for many years now, and the experience of the legal profession with such systems may be instructive. The most important question about information retrieval systems is whether they find the things they are supposed to find. Collen and Flagle note that the very small pilot data base they used was not big enough to evaluate the retrieval performance of the MEDIS system. The leading study of the retrieval effectiveness of full-text systems was conducted by Blair and Maron 2 and was cited but not discussed by Collen and Flagel. Blair and Maron found that a full-text system they studied retrieved no more