The Experts below are selected from a list of 337914 Experts worldwide ranked by ideXlab platform
Furman S Mcdonald - One of the best experts on this subject based on the ideXlab platform.
-
relationship of electronic Medical Knowledge resource use and practice characteristics with internal medicine maintenance of certification examination scores
Journal of General Internal Medicine, 2012Co-Authors: Darcy A Reed, Rebecca S Lipner, Eric S Holmboe, Colin P West, Andrew J Halvorsen, Carola Jacobs, Furman S McdonaldAbstract:BACKGROUND Maintenance of certification examination performance is associated with quality of care. We aimed to examine relationships between electronic Medical Knowledge resource use, practice characteristics and examination scores among physicians recertifying in internal medicine.
-
factors associated with Medical Knowledge acquisition during internal medicine residency
Journal of General Internal Medicine, 2007Co-Authors: Furman S Mcdonald, Scott L Zeger, Joseph C KolarsAbstract:BACKGROUND Knowledge acquisition is a goal of residency and is measurable by in-training exams. Little is known about factors associated with Medical Knowledge acquisition.
William R Hogan - One of the best experts on this subject based on the ideXlab platform.
-
combine factual Medical Knowledge and distributed word representation to improve clinical named entity recognition
American Medical Informatics Association Annual Symposium, 2018Co-Authors: Yonghui Wu, Xi Yang, Jiang Bian, Hua Xu, William R HoganAbstract:: There has been an increasing interest in developing deep learning methods to recognize clinical concepts from narrative clinical text. Recently, several studies have reported that Recurrent Neural Networks (RNNs) outperformed traditional machine learning methods such as Conditional Random Fields (CRFs). Deep learning-based Named Entity Recognition (NER) systems often use statistical language models to learn word embeddings from unlabeled corpora. However, current word embedding methods have limitations to learn decent representations for low-frequency words. Medicine is a Knowledge-extensive domain; existing Medical Knowledge has the potential to improve feature representations for less frequent yet important words. However, it is still not clear how existing Medical Knowledge can help deep learning models in clinical NER tasks. In this study, we integrated Medical Knowledge from the Unified Medical Language System with word embeddings trained from an unlabeled clinical corpus in RNNs for detection of problems, treatments and lab tests. We examined three different ways to generate Medical Knowledge features, including a dictionary lookup program, the KnowledgeMap system, and the MedLEE system. We also compared representing Medical Knowledge as one-hot vectors versus representing Medical Knowledge as embedding layers. The evaluation results showed that the RNN with Medical Knowledge as embedding layers achieved new state-of-the-art performance (a strict F1 score of 86.21% and a relaxed F1 score of 92.80%) on the 2010 i2b2 corpus, outperforming an RNN with only word embeddings and RNNs with Medical Knowledge as one-hot vectors. This study demonstrated an efficient way of integrating Medical Knowledge with distributed word representations for clinical NER.
-
combine factual Medical Knowledge and distributed word representation to improve clinical named entity recognition
American Medical Informatics Association Annual Symposium, 2018Co-Authors: Xi Yang, Jiang Bian, Yi Guo, William R HoganAbstract:There has been an increasing interest in developing deep learning methods to recognize clinical concepts from narrative clinical text. Recently, several studies have reported that Recurrent Neural Networks (RNNs) outperformed traditional machine learning methods such as Conditional Random Fields (CRFs). Deep learning-based Named Entity Recognition (NER) systems often use statistical language models to learn word embeddings from unlabeled corpora. However, current word embedding methods have limitations to learn decent representations for low-frequency words. Medicine is a Knowledge-extensive domain; existing Medical Knowledge has the potential to improve feature representations for less frequent yet important words. However, it is still not clear how existing Medical Knowledge can help deep learning models in clinical NER tasks. In this study, we integrated Medical Knowledge from the Unified Medical Language System with word embeddings trained from an unlabeled clinical corpus in RNNs for detection of problems, treatments and lab tests. We examined three different ways to generate Medical Knowledge features, including a dictionary lookup program, the KnowledgeMap system, and the MedLEE system. We also compared representing Medical Knowledge as one-hot vectors versus representing Medical Knowledge as embedding layers. The evaluation results showed that the RNN with Medical Knowledge as embedding layers achieved new state-of-the-art performance (a strict F1 score of 86.21% and a relaxed F1 score of 92.80%) on the 2010 i2b2 corpus, outperforming an RNN with only word embeddings and RNNs with Medical Knowledge as one-hot vectors. This study demonstrated an efficient way of integrating Medical Knowledge with distributed word representations for clinical NER.
Joseph C Kolars - One of the best experts on this subject based on the ideXlab platform.
-
quality of life burnout educational debt and Medical Knowledge among internal medicine residents
JAMA, 2011Co-Authors: Colin Patrick West, Tait D. Shanafelt, Joseph C KolarsAbstract:Context Physician distress is common and has been associated with negative effects on patient care. However, factors associated with resident distress and well-being have not been well described at a national level. Objectives To measure well-being in a national sample of internal medicine residents and to evaluate relationships with demographics, educational debt, and Medical Knowledge. Design, Setting, and Participants Study of internal medicine residents using data collected on 2008 and 2009 Internal Medicine In-Training Examination (IM-ITE) scores and the 2008 IM-ITE survey. Participants were 16 394 residents, representing 74.1% of all eligible US internal medicine residents in the 2008-2009 academic year. This total included 7743 US Medical graduates and 8571 international Medical graduates. Main Outcome Measures Quality of life (QOL) and symptoms of burnout were assessed, as were year of training, sex, Medical school location, educational debt, and IM-ITE score reported as percentage of correct responses. Results Quality of life was rated “as bad as it can be” or “somewhat bad” by 2402 of 16 187 responding residents (14.8%). Overall burnout and high levels of emotional exhaustion and depersonalization were reported by 8343 of 16 192 (51.5%), 7394 of 16 154 (45.8%), and 4541 of 15 737 (28.9%) responding residents, respectively. In multivariable models, burnout was less common among international Medical graduates than among US Medical graduates (45.1% vs 58.7%; odds ratio, 0.70 [99% CI, 0.63-0.77]; P $200 000 relative to no debt). Residents reporting QOL “as bad as it can be” and emotional exhaustion symptoms daily had mean IM-ITE scores 2.7 points (99% CI, 1.2-4.3; P Conclusions In this national study of internal medicine residents, suboptimal QOL and symptoms of burnout were common. Symptoms of burnout were associated with higher debt and were less frequent among international Medical graduates. Low QOL, emotional exhaustion, and educational debt were associated with lower IM-ITE scores.
-
factors associated with Medical Knowledge acquisition during internal medicine residency
Journal of General Internal Medicine, 2007Co-Authors: Furman S Mcdonald, Scott L Zeger, Joseph C KolarsAbstract:BACKGROUND Knowledge acquisition is a goal of residency and is measurable by in-training exams. Little is known about factors associated with Medical Knowledge acquisition.
Zhou Xiao - One of the best experts on this subject based on the ideXlab platform.
-
Medical Knowledge acquisition an ontology based approach
Computer Science, 2003Co-Authors: Zhou XiaoAbstract:In this paper, we introduce an ontology-mediated method for Medical Knowledge acquisition and analysis. Using the method we establish an ontological structure and ontologies for the Medical Knowledge Base (or NKIMed). To check the consistency of the acquired Knowledge, we use a set of medicine-specific axioms. These axioms are also used in Knowledge inference, and interconnection between different Medical concepts. Finally, two applications of NKIMed, i.e. intelligent teaching systems and speech diagnosis are illustrated.
Xi Yang - One of the best experts on this subject based on the ideXlab platform.
-
combine factual Medical Knowledge and distributed word representation to improve clinical named entity recognition
American Medical Informatics Association Annual Symposium, 2018Co-Authors: Yonghui Wu, Xi Yang, Jiang Bian, Hua Xu, William R HoganAbstract:: There has been an increasing interest in developing deep learning methods to recognize clinical concepts from narrative clinical text. Recently, several studies have reported that Recurrent Neural Networks (RNNs) outperformed traditional machine learning methods such as Conditional Random Fields (CRFs). Deep learning-based Named Entity Recognition (NER) systems often use statistical language models to learn word embeddings from unlabeled corpora. However, current word embedding methods have limitations to learn decent representations for low-frequency words. Medicine is a Knowledge-extensive domain; existing Medical Knowledge has the potential to improve feature representations for less frequent yet important words. However, it is still not clear how existing Medical Knowledge can help deep learning models in clinical NER tasks. In this study, we integrated Medical Knowledge from the Unified Medical Language System with word embeddings trained from an unlabeled clinical corpus in RNNs for detection of problems, treatments and lab tests. We examined three different ways to generate Medical Knowledge features, including a dictionary lookup program, the KnowledgeMap system, and the MedLEE system. We also compared representing Medical Knowledge as one-hot vectors versus representing Medical Knowledge as embedding layers. The evaluation results showed that the RNN with Medical Knowledge as embedding layers achieved new state-of-the-art performance (a strict F1 score of 86.21% and a relaxed F1 score of 92.80%) on the 2010 i2b2 corpus, outperforming an RNN with only word embeddings and RNNs with Medical Knowledge as one-hot vectors. This study demonstrated an efficient way of integrating Medical Knowledge with distributed word representations for clinical NER.
-
combine factual Medical Knowledge and distributed word representation to improve clinical named entity recognition
American Medical Informatics Association Annual Symposium, 2018Co-Authors: Xi Yang, Jiang Bian, Yi Guo, William R HoganAbstract:There has been an increasing interest in developing deep learning methods to recognize clinical concepts from narrative clinical text. Recently, several studies have reported that Recurrent Neural Networks (RNNs) outperformed traditional machine learning methods such as Conditional Random Fields (CRFs). Deep learning-based Named Entity Recognition (NER) systems often use statistical language models to learn word embeddings from unlabeled corpora. However, current word embedding methods have limitations to learn decent representations for low-frequency words. Medicine is a Knowledge-extensive domain; existing Medical Knowledge has the potential to improve feature representations for less frequent yet important words. However, it is still not clear how existing Medical Knowledge can help deep learning models in clinical NER tasks. In this study, we integrated Medical Knowledge from the Unified Medical Language System with word embeddings trained from an unlabeled clinical corpus in RNNs for detection of problems, treatments and lab tests. We examined three different ways to generate Medical Knowledge features, including a dictionary lookup program, the KnowledgeMap system, and the MedLEE system. We also compared representing Medical Knowledge as one-hot vectors versus representing Medical Knowledge as embedding layers. The evaluation results showed that the RNN with Medical Knowledge as embedding layers achieved new state-of-the-art performance (a strict F1 score of 86.21% and a relaxed F1 score of 92.80%) on the 2010 i2b2 corpus, outperforming an RNN with only word embeddings and RNNs with Medical Knowledge as one-hot vectors. This study demonstrated an efficient way of integrating Medical Knowledge with distributed word representations for clinical NER.
