11/10/2023 0 Comments Deep synonym medical![]() Sadly, creating inventories at every US healthcare institution is not feasible, especially without fully automated methods which do not exist.ĭeep data refers to high-quality, complete, and relevant data with an internal structure that may be large-scale 37, 38. These inventories are sufficient for institution-specific tasks, but have been inadequate for cross-institutional (interoperable) tasks, because abbreviations vary substantially based on medical specialty and setting 23, 24, 25, 26. Due to this limitation, several institutions have engineered their own smaller, more clinically-oriented sense inventories 30, 31, 32, 33, 34, 35, 36. Large sense inventories can be publicly obtained online (e.g., Unified Medical Language System, /research/umls) but they can be incomplete 13, 19, 27, 28, because they were generated using biological research corpora such as research papers, not clinical corpora such as electronic health records 13, 29. ![]() Recognition, disambiguation, and expansion of abbreviations relies on sense inventories, defined as databases of abbreviations and their meanings or senses. Furthermore, recognizing, disambiguating, and expanding abbreviations can help physicians, nurses, caregivers, and patients understand them, which studies have shown prevents medically-dangerous misinterpretation 22, 23, 24, 25, 26. As such, recognizing, disambiguating, and expanding abbreviations is central to clinical NLP, and even small advancements would improve performance and practical application 11, 12, 13, 14, 15, 16, 17, 18, 19. Abbreviations constitute 30–50% of the words in clinical text, such as doctor’s notes 20, compared to <1% in general text, such as news media 21. ![]() Since then, pre-trained transformer architectures have become mainstream for language tasks involving contextual long-distance dependencies, and have been incorporated into commercial services such as Google Search 9 and Amazon Alexa 10.ĭespite these recent advancements, clinical abbreviations and acronyms (hereafter, ‘abbreviations’) persistently impede NLP performance and practical application in health and healthcare 11, 12, 13, 14, 15, 16, 17, 18, 19. These breakthroughs have empowered researchers to build generalizable language models and apply them to achieve superior accuracy on subsequent downstream tasks 8. In the past few years, artificial intelligence breakthroughs using pre-trained transformer architectures have revolutionized NLP 7. NLP translates free text and speech into standardized data 3, which can help clinicians make decisions 4, predict health outcomes 5, prevent adverse events 6, and improve quality-of-care 1, 2. Natural language processing (NLP) is becoming essential to health and healthcare 1, 2. This allows for cross-institutional natural language processing, which previous inventories did not support. The multiple sources and high coverage support application in varied specialties and settings. To our knowledge, the Meta-Inventory is the most complete compilation of medical abbreviations and acronyms in American English to-date. The Meta-Inventory demonstrated high completeness or coverage of abbreviations and senses in new clinical text, a substantial improvement over the next largest repository (6–14% increase in abbreviation coverage 28–52% increase in sense coverage). Additional features include semi-automated quality control to remove errors. Automated cross-mapping of synonymous records using state-of-the-art machine learning reduced redundancy, which simplifies future application. A systematic harmonization of eight source inventories across multiple healthcare specialties and settings identified 104,057 abbreviations with 170,426 corresponding senses. To support recognition, disambiguation, and expansion, we present the Medical Abbreviation and Acronym Meta-Inventory, a deep database of medical abbreviations. The recognition, disambiguation, and expansion of medical abbreviations and acronyms is of upmost importance to prevent medically-dangerous misinterpretation in natural language processing.
0 Comments
Leave a Reply. |
AuthorWrite something about yourself. No need to be fancy, just an overview. ArchivesCategories |