Article
Chemistry, Multidisciplinary
Tanmoy Paul, Humayera Islam, Nitesh Singh, Yaswitha Jampani, Teja Venkat Pavan Kotapati, Preethi Aishwarya Tautam, Md Kamruz Zaman Rana, Vasanthi Mandhadi, Vishakha Sharma, Michael Barnes, Richard D. Hammer, Abu Saleh Mohammad Mosa
Summary: The de-identification of clinical reports is crucial for patient confidentiality protection. This study explores the utility of various features in a conditional-random-field-based named entity recognition model through annotating a large volume of radiology reports and building NER models. The results indicate that n-gram, prefix-suffix, word embedding, and word shape are the best-performing features.
APPLIED SCIENCES-BASEL
(2022)
Article
Immunology
Constantina Boikos, Mahrukh Imran, Simon de Lusignan, Justin R. Ortiz, Peter A. Patriarca, James A. Mansi
Summary: Real-world evidence (RWE) plays an increasingly important role in public health and healthcare decisions globally. An Integrated Dataset has been created, combining primary care electronic medical records with pharmacy and medical claims data for over 123 million US patients since 2014. This study describes the components of the Integrated Dataset and evaluates its representativeness to the US population, as well as its potential for evaluating influenza vaccine effectiveness.
Article
Biochemical Research Methods
Aranka Steyaert, Pieter Audenaert, Jan Fostier
Summary: In de novo genome assembly, accurate determination of node and arc multiplicities in a de Bruijn graph is crucial for improving assembly quality and contiguity. A Conditional Random Field (CRF) model is proposed to model the de Bruijn graph and read coverage information. Loopy Belief Propagation (LBP) is applied for approximate inference to improve multiplicity assignment accuracy. The order of message passing in LBP greatly affects its convergence speed, and an empirical evaluation of different message passing schemes is presented.
IEEE-ACM TRANSACTIONS ON COMPUTATIONAL BIOLOGY AND BIOINFORMATICS
(2023)
Article
Health Care Sciences & Services
Sungrim Moon, Luke A. Carlson, Ethan D. Moser, Bhavani Singh Agnikula Kshatriya, Carin Y. Smith, Walter A. Rocca, Liliana Gazzuola Rocca, Suzette J. Bielinski, Hongfang Liu, Nicholas B. Larson
Summary: This study compares the gynecological surgical history extracted from an EHR using natural language processing techniques with manually curated surgical history information. The study finds that missing or inaccurately reported and recorded information contributes to the misclassification of surgical status.
JOURNAL OF MEDICAL INTERNET RESEARCH
(2022)
Article
Health Care Sciences & Services
Jelle Keuper, Ronald Batenburg, Lilian van Tuyl, Robert Verheij
Summary: The provision of web-based access to medical records by general practices in the Netherlands has been mainly positive, despite the increase in patient consultations and administrative actions. Earlier adoption of this service was associated with a more positive attitude and experienced effects.
JOURNAL OF MEDICAL INTERNET RESEARCH
(2023)
Article
Physics, Multidisciplinary
Ursula Torres Parejo, Jesus Roque Campana, Maria Amparo Vila, Miguel Delgado
Summary: The study aims to allow visual exploration of information in medical databases using a novel mathematical method and multi-term tag clouds as content representation tools. Through a survey of a Spanish database with 24,481 records, the effectiveness of tag clouds as information retrieval tools was evaluated. The research demonstrates that the automatically generated visual interface captures the semantics of information and facilitates access to medical records.
Article
Computer Science, Information Systems
Sheng-Hu Tian, Yu Chen, Tao-Zhen Bai
Summary: The main purpose of this study is to identify the key factors affecting the vertical integration of electronic health records in medical consortia and clarify their impact mechanism, providing reference for improving relevant policies. The fuzzy-DEMATEL-ASIM method was used to establish an index system of influencing factors and reveal the influence relationship and action mechanism among the key factors using a multi-layer hierarchical influence structure model. The results showed that there are 32 factors influencing the vertical integration, 17 of which are key factors. These key factors should be prioritized for improvement and optimization to promote integration projects, and future policies should focus on them.
INTERNATIONAL JOURNAL OF MEDICAL INFORMATICS
(2023)
Review
Psychiatry
Yvonne Zurynski, Louise A. Ellis, Huong Ly Tong, Liliana Laranjo, Robyn Clay-Williams, Luke Testa, Isabelle Meulenbroeks, Charmaine Turton, Grant Sara
Summary: The successful implementation of EMRs in mental health settings is influenced by financial, technical, and organizational factors, as well as clinicians' perception of the appropriateness and acceptability of the system. Clinicians value EMRs that do not disrupt workflow, improve documentation completeness and accuracy, and support quality of care.
JMIR MENTAL HEALTH
(2021)
Article
Environmental Sciences
Yingying Kong, Biyuan Yan, Yanjuan Liu, Henry Leung, Xiangyang Peng
Summary: The study proposes a classification model based on feature-level fusion of optical images and Synthetic Aperture Radar (SAR), improving feature identification capabilities. By introducing feature importance to enhance classification accuracy, the dataset combining the two shows significant improvements over using optical or polarized SAR image features alone. The random forest-importance_ conditional random fields (RF-Im_CRF) model developed in this paper achieved the best overall accuracy and Kappa coefficient among the four classification models used, validating the effectiveness of the method.
Article
Computer Science, Interdisciplinary Applications
Fengdong Chi, Pierre Breul, Claudio Carvajal, Laurent Peyras
Summary: This study performs a stochastic analysis of seepage through an embankment dam by using and comparing three types of random fields (RFs). The results show that different RFs have different impacts on the seepage results, with significant effects on flow velocity, moderate effects on hydraulic gradient, and minor effects on pore pressure.
COMPUTERS AND GEOTECHNICS
(2023)
Article
Computer Science, Information Systems
Zfania Tom Korach, Stephen Gradwohl, Amanda Messinger, Kelly Bookman, Kevin Cohen, Li Zhou, Foster Goss
Summary: The study compared knowledge-based and unsupervised statistical methods for ranking clinical information in Emergency Department patients with chest or back pain complaints. The results showed that the unsupervised statistical method outperformed the knowledge-based ranking for problems, but underperformed for medications.
INTERNATIONAL JOURNAL OF MEDICAL INFORMATICS
(2021)
Article
Medical Informatics
Peng Wang, Yong Li, Liang Yang, Simin Li, Linfeng Li, Zehan Zhao, Shaopei Long, Fei Wang, Hongqian Wang, Ying Li, Chengliang Wang
Summary: This paper proposes a method to overcome the difficulties of overfitting and a lack of training data for deep neural networks to enable Chinese protected health information deidentification. The method merges TinyBERT as a text feature extraction module and the conditional random field method as a prediction module for deidentifying protected health information in Chinese medical electronic health records.
JMIR MEDICAL INFORMATICS
(2022)
Article
Health Care Sciences & Services
Louis Mercorelli, Harrison Nguyen, Nicole Gartell, Martyn Brookes, Jonathan Morris, Charmaine S. Tam
Summary: Clinical free-text data is a vast and untapped source of rich information. However, the lack of transparency in existing de-identification software tools makes it difficult to assess the potential risks associated with releasing de-identified clinical free-text data. This case study describes the development of a framework for releasing de-identified clinical free-text data and ensuring data safety through algorithm design and manual review.
AUSTRALIAN HEALTH REVIEW
(2022)
Article
Computer Science, Interdisciplinary Applications
Xiangfeng Guo, Daniel Dias, Claudio Carvajal, Laurent Peyras, Pierre Breul
Summary: The study uses different types of RF for soil spatial modeling and investigates their effects on dam reliability. The results show that using a simple RF is sufficient for satisfactory results in well-controlled dam constructions, while non-typical RFs may require more accurate data.
ENGINEERING WITH COMPUTERS
(2022)
Article
Computer Science, Information Systems
Liran Politi, Shlomi Codish, Iftach Sagy, Lior Fink
Summary: This study fills the research gap in understanding how the use of one health information system changes after the implementation of another system. The findings reveal that following the implementation of a local electronic medical record system, physicians accessed the health information exchange system less frequently, but the volume and duration of system use increased. These results suggest that previous research has underestimated the explanatory value of cost-benefit approach and the occurrence of complementarity in system use due to the reliance on a narrow definition of use.
EUROPEAN JOURNAL OF INFORMATION SYSTEMS
(2022)