AMIA ... Annual Symposium Proceedings / AMIA Symposium. AMIA Symposium[JOURNAL]
Sun
200 papers
Data Modernization in Action: Synthesizing Pioneering Informatics Projects in Public Health and Data Modernization Stories from Public Health Agencies.
Kim C, Schmit AD, Solarz S … +1 more , Rajamani S
AMIA Annu Symp Proc · 2024 · PMID 41726523
With data considered as the 'oxygen' of public health, the Data Modernization Initiative (DMI) to enhance the public health data and information infrastructure is critical. The DMI Stories from the Field features data mo...
With data considered as the 'oxygen' of public health, the Data Modernization Initiative (DMI) to enhance the public health data and information infrastructure is critical. The DMI Stories from the Field features data modernization from public health agencies to highlight success/progress. These stories (n=241) were analyzed, with outbreak response, information systems capacity, epidemiology/laboratory capacity being some of the common topics. A total of 199 codes across DMI stories were organized into 7 themes and the top 3 codes were communication, collaboration and public health agencies. Key takeaways and next steps were identified and validated with expert input across people, product, process and partnership categories and people factor was critical along with funding/sustainability. Ongoing DMI stories and future studies for evaluating impact are recommended. DMI stories are a great option to communicate the projects and impact of DMI to a larger public audience and garner support for this vital endeavor.
Recruitment, Retention, and Participant Diversity in the Context of a Heart Failure Self-Care Digital Intervention Decentralized Clinical Trial.
West E, Rangel A, Zeng J … +3 more , Im C, Shah F, Radhakrishnan K
AMIA Annu Symp Proc · 2024 · PMID 41726522
BACKGROUND: Low participant accrual is a persistent concern in physiological disease intervention trials, inflating costs and jeopardizing the timeliness and validity of findings. Investigators are increasingly adopting...
BACKGROUND: Low participant accrual is a persistent concern in physiological disease intervention trials, inflating costs and jeopardizing the timeliness and validity of findings. Investigators are increasingly adopting decentralized methods to facilitate participation. OBJECTIVE: To add to the recruitment evidence base by describing the performance of direct and remote recruitment strategies in a decentralized randomized controlled trial of a digital intervention to improve heart failure self-care behaviors. METHODS: We conducted a descriptive analysis of referral, enrollment, and retention rates; cost; and sociodemographic diversity of participants across six recruitment streams. Data were collected from Sept 30, 2022 to June 30, 2025. RESULTS: Decentralized recruitment channels generated 97.5% of enrollments and achieved varying success with respect to sample representativeness. Enrollment rates progressed in accordance with proposed timelines. Retention at 6 months was 82.6%. CONCLUSIONS: Decentralized recruitment strategies are feasible, cost effective, and conducive to achieving enrollment targets.
Mining Social Media Data for Influenza Vaccine Effectiveness Using a Large Language Model and Chain-of-Thought Prompting.
Xu D, García GL, O'Connor K … +5 more , Holston H, Klein AZ, Amaro IF, Scotch M, Gonzalez-Hernandez G
AMIA Annu Symp Proc · 2024 · PMID 41726521
Influenza vaccine effectiveness (VE) estimation plays a critical role in public health decision-making by quantifying the real-world impact of vaccination campaigns and guiding policy adjustments. Current approaches to V...
Influenza vaccine effectiveness (VE) estimation plays a critical role in public health decision-making by quantifying the real-world impact of vaccination campaigns and guiding policy adjustments. Current approaches to VE estimation are constrained by limited population representation, selection bias, and delayed reporting. To address some of these gaps, we propose leveraging large language models (LLMs) with few-shot chain-of-thought (CoT) prompting to mine social media data for real-time influenza VE estimation. We annotated over 4,000 tweets from the 2020-2021 flu season using structured guidelines, achieving high inter-annotator agreement. Our best prompting strategy achieves F scores above 87% for identifying influenza vaccination status and test outcomes, outperforming traditional supervised fine-tuning methods by large margins. These findings indicate that LLM-based prompting approaches effectively identify relevant social media information for influenza VE estimation, offering a valuable real-time surveillance tool that complements traditional epidemiological methods.
Detecting Reference Errors in Scientific Literature with Large Language Models.
Zhang TM, Abernethy NF
AMIA Annu Symp Proc · 2024 · PMID 41726520
Reference errors, such as citation and quotation errors, are common in scientific papers. Such errors can result in the propagation of inaccurate information, but are difficult and time-consuming to detect, posing a sign...
Reference errors, such as citation and quotation errors, are common in scientific papers. Such errors can result in the propagation of inaccurate information, but are difficult and time-consuming to detect, posing a significant threat to the integrity of scientific literature. To support automatic detection of reference errors, we evaluated the ability of large language models in OpenAI's GPT family to detect quotation errors. Specifically, we prepared an expert-annotated, general-domain dataset of statement-reference pairs from journal articles, one-third of which is in biomedicine. Large language models were evaluated in different settings with varying amounts of reference information provided by retrieval augmentation. Results showed that large language models are able to detect erroneous citations with limited context and without fine-tuning. This study contributes to the growing literature that seeks to utilize artificial intelligence to assist in the writing, reviewing, and publishing of scientific papers as well as grounding of language model responses.
Opportunistic Screening for Pancreatic Cancer using Computed Tomography Imaging and Radiology Reports.
Le D, Correa-Medero R, Tariq A … +3 more , Patel B, Yano M, Banerjee I
AMIA Annu Symp Proc · 2024 · PMID 41726519
Pancreatic ductal adenocarcinoma (PDAC) is a highly aggressive cancer, with most cases diagnosed at stage IV and a five-year overall survival rate below 5%. Early detection and prognosis modeling are crucial for improvin...
Pancreatic ductal adenocarcinoma (PDAC) is a highly aggressive cancer, with most cases diagnosed at stage IV and a five-year overall survival rate below 5%. Early detection and prognosis modeling are crucial for improving patient outcomes and guiding early intervention strategies. In this study, we implemented and evaluated deep learning fusion models that integrate radiology reports and CT imaging to predict PDAC risk. The DeepSurv model achieved a concordance index (C-index) of 0.6773 (95% CI: 0.6484, 0.7061) and 0.6596 (95% CI: 0.6260, 0.6937) on the internal and external dataset, respectively, for 5-year survival risk estimation. Kaplan-Meier analysis demonstrated significant separation (p<0.0001) between the low and high risk groups predicted by the fusion model. These findings highlight the potential of deep learning-based survival models in leveraging clinical and imaging data for pancreatic cancer.
ProtoBERT-LoRA: Parameter-Efficient Prototypical Finetuning for Immunotherapy Study Identification.
Zhang S, Ding X, Ding K … +6 more , Zhang J, Galinsky K, Wang M, Mayers RP, Wang Z, Kharrazi H
AMIA Annu Symp Proc · 2024 · PMID 41726518
Identifying immune checkpoint inhibitor (ICI) studies in genomic repositories like Gene Expression Omnibus (GEO) is vital for cancer research yet remains challenging due to semantic ambiguity, extreme class imbalance, an...
Identifying immune checkpoint inhibitor (ICI) studies in genomic repositories like Gene Expression Omnibus (GEO) is vital for cancer research yet remains challenging due to semantic ambiguity, extreme class imbalance, and limited labeled data in low-resource settings. We present ProtoBERT-LoRA, a hybrid framework that combines PubMedBERT with prototypical networks and Low-Rank Adaptation (LoRA) for efficient fine-tuning. The model enforces class-separable embeddings via episodic prototype training while preserving biomedical domain knowledge. Our dataset was divided as: Training (20 positive, 20 negative), Prototype Set (10 positive, 10 negative), Validation (20 positive, 200 negative), and Test (71 positive, 765 negative). Evaluated on test dataset, ProtoBERT-LoRA achieved F1-score of 0.624 (precision: 0.481, recall: 0.887), outperforming the rule-based system, machine learning baselines and finetuned PubMedBERT. Application to 44,287 unlabeled studies reduced manual review efforts by 82%. Ablation studies confirmed that combining prototypes with LoRA improved performance by 29% over stand-alone LoRA.
Generating Counterfactual Patient Timelines from Real-World Data.
Akagi Y, Seki T, Takiguchi T … +3 more , Ito H, Kawazoe Y, Ohe K
AMIA Annu Symp Proc · 2024 · PMID 41726517
Counterfactual simulation-exploring hypothetical consequences under alternative clinical scenarios-holds promise for transformative applications such as personalized medicine and in-silico trials. However, it remains cha...
Counterfactual simulation-exploring hypothetical consequences under alternative clinical scenarios-holds promise for transformative applications such as personalized medicine and in-silico trials. However, it remains challenging due to methodological limitations. Here, we show that an autoregressive generative model, trained on real-world data from over 300,000 patients and 400 million patient timeline entries, can generate clinically plausible counterfactual trajectories. As a validation task, we applied the model to patients hospitalized with COVID-19 in 2023, modifying age, serum C-reactive protein (CRP), and serum creatinine to simulate 7-day outcomes. Increased in-hospital mortality was observed in counterfactual simulations with older age, elevated CRP, and elevated serum creatinine. Remdesivir prescriptions increased in simulations with higher CRP values and decreased in those with impaired kidney function. These counterfactual trajectories reproduced known clinical patterns. These findings suggest that autoregressive generative models trained on real-world data in a self-supervised manner can establish a foundation for counterfactual clinical simulation.
Enhancing Long-Term Care Efficiency: Embedded LLMs for Clinical Report Summarization and Caregiver Support.
Michelet A, Manzo G, Ritz A … +3 more , Delgado P, Celi LA, Schumacher MI
AMIA Annu Symp Proc · 2024 · PMID 41726516
Long-term care facilities face a critical shortage of nursing staff and an increasing administrative burden, reducing time for direct patient care. Generative artificial intelligence offers a promising solution to automa...
Long-term care facilities face a critical shortage of nursing staff and an increasing administrative burden, reducing time for direct patient care. Generative artificial intelligence offers a promising solution to automate administrative tasks and support caregivers. This paper evaluates the relevance of using a fine-tuned large language model (LLM) to address these challenges. Interviews with healthcare professionals identified key needs, leading to the selection of two use cases: caregiver-patient communication assistance and medical record summarization. To comply with privacy and security constraints, the model was deployed in an embedded scenario. Performance evaluations showed significant improvements in BLEU and ROUGE metrics for both use cases, demonstrating enhanced accuracy. This study demonstrates the feasibility of leveraging LLMs to streamline workflows, reduce administrative strain, and improve operational efficiency. This work highlights the potential for broader AI applications in long-term care, paving the way for better working conditions for caregivers and improved patient care quality.
Ontology-based Semantic Similarity Measures for Clustering Medical Concepts in Drug Safety.
Painter JL, Haguinet F, Powell GE … +1 more , Bate A
AMIA Annu Symp Proc · 2024 · PMID 41726515
Semantic similarity measures (SSMs) are widely used in biomedical research but remain underutilized in pharmacovigilance. This study evaluates six ontology-based SSMs for clustering MedDRA Preferred Terms (PTs) in drug s...
Semantic similarity measures (SSMs) are widely used in biomedical research but remain underutilized in pharmacovigilance. This study evaluates six ontology-based SSMs for clustering MedDRA Preferred Terms (PTs) in drug safety data. Using the Unified Medical Language System (UMLS), we assess each method's ability to group PTs around medically meaningful centroids. A high-throughput framework was developed with a Java API and Python/R interfaces support large-scale similarity computations. Results show that while path-based methods perform moderately with F1 scores of 0.36 for WUPALMER and 0.28 for LCH, intrinsic information content (IC)-based measures, especially INTRINSIC_LIN and SOKAL, consistently yield better clustering accuracy (F1 Score of 0.403). Validated against expert review and standard MedDRA queries (SMQs), our findings highlight the promise of IC-based SSMs in enhancing pharmacovigilance workflows by improving early signal detection and reducing manual review.
Development of an Early-Phase Local Model for Pandemics Using Public Health Data: Application to the COVID-19 Pandemic.
Shi J, Morris JS, Rubin D … +1 more , Huang J
AMIA Annu Symp Proc · 2024 · PMID 41726514
The emergence of novel infectious pathogens challenges early-phase modeling of disease transmission due to limited, low-quality data and an incomplete understanding of the pathogen. Additionally, regional variations in o...
The emergence of novel infectious pathogens challenges early-phase modeling of disease transmission due to limited, low-quality data and an incomplete understanding of the pathogen. Additionally, regional variations in outbreaks necessitate models that incorporate local dynamics. We present an early-phase local model that leverages constrained public health data, primarily infection counts and aggregated regional characteristics, to study disease transmission dynamics. To address data limitations and potential model misspecifications, we incorporate a quasi-likelihood approach with a flexible error term. Furthermore, we introduce an online estimator that enables real-time data updates, supported by an iterative algorithm for parameter estimation. We applied this method to early COVID-19 data, analyzing infection counts and county-level risk factors from more than 800 U.S. counties to predict disease spread and assess the impact of social behavior, demographics, and vaccination coverage on disease transmission. This framework improves early outbreak analysis and informs local pandemic response under suboptimal data conditions.
Exploring and Comparing Existing Natural Product Databases Towards Whole Person Health Research.
Chen X, Song M, Hou Y … +5 more , Rizvi RF, Bishop JR, Ranallo PA, Hoye TR, Zhang R
AMIA Annu Symp Proc · 2024 · PMID 41726513
Natural products are essential in drug discovery, chemical biology, and medicinal chemistry. Despite their widespread use, NP data remains fragmented across various databases, limiting their utility for whole person heal...
Natural products are essential in drug discovery, chemical biology, and medicinal chemistry. Despite their widespread use, NP data remains fragmented across various databases, limiting their utility for whole person health research, which requires comprehensive, interoperable resources. This study explores and compares three major NP databases: COCONUT, NP-MRD, and GSRS, assessing their scope, structural representation, metadata completeness, and accessibility. COCONUT provides extensive chemical diversity, NP-MRD emphasizes spectral and physical property data, and GSRS focuses on regulatory classification. Despite their strengths, overlap between databases is moderate to small, and significant gaps remain in integrating medical and pharmaceutical information. Improved interoperability and harmonization are needed to support advanced computational models for whole person health. Our findings highlight critical gaps and opportunities to enhance NP database integration, laying the groundwork for developing comprehensive resources that better support data-driven investigations of natural products.
Towards Interpretable, Sequential Multiple Instance Learning: An Application to Clinical Imaging.
Luo X, Wang HS, Li ML
AMIA Annu Symp Proc · 2024 · PMID 41726512
This work introduces the Sequential Multiple Instance Learning (SMIL) framework, addressing the challenge of interpreting sequential, variable-length sequences of medical images with a single diagnostic label. Diverging...
This work introduces the Sequential Multiple Instance Learning (SMIL) framework, addressing the challenge of interpreting sequential, variable-length sequences of medical images with a single diagnostic label. Diverging from traditional MIL approaches that treat image sequences as unordered sets, SMIL systematically integrates the sequential nature of clinical imaging. We develop a bidirectional Transformer architecture, BiSMIL, that optimizes for both early and final prediction accuracies through a novel training procedure to balance diagnostic accuracy with operational efficiency. We evaluated BiSMIL on three medical image datasets to demonstrate that it simultaneously achieves state-of-the-art final accuracy and superior performance in early prediction accuracy, requiring 30-50% fewer images for a similar level of performance compared to existing models. Additionally, we introduce SMILU, an interpretable uncertainty metric that outperforms traditional metrics in identifying challenging instances.
Supporting Patients in Managing Electronic Health Records and Biospecimens Consent for Research: Insights from a Mixed-Methods Usability Evaluation of the iAGREE Portal.
Hu D, Lu X, Chen Y … +6 more , Keller M, Nguyen AT, Le V, Kuo TT, Ohno-Machado L, Zheng K
AMIA Annu Symp Proc · 2024 · PMID 41726511
De-identified health data are frequently used in research. As AI advances heighten the risk of re-identification, it is important to respond to concerns about transparency, data privacy, and patient preferences. However,...
De-identified health data are frequently used in research. As AI advances heighten the risk of re-identification, it is important to respond to concerns about transparency, data privacy, and patient preferences. However, few practical and user-friendly solutions exist. We developed iAGREE, a patient-centered electronic consent management portal that allows patients to set granular preferences for sharing electronic health records and biospecimens with researchers. To refine the iAGREE portal, we conducted a mixed-methods usability evaluation with 40 participants from three U.S. health systems. Our results show that the portal received highly positive usability feedback. Moreover, participants identified areas for improvement, suggested actionable enhancements, and proposed additional features to better support informed granular consent while reducing patient burden. Insights from this study may inform further improvements to iAGREE and provide practical guidance for designing patient-centered consent management tools.
Comparing Patient-level Social Drivers of Health from Health Surveys and Electronic Health Records for Patients with Comorbid Hypertension and Uncontrolled Diabetes.
Sang S, Silva SG, Spratt SE … +5 more , Palipana AK, Matos LA, Fitzpatrick M, Crowley MJ, Shaw RJ
AMIA Annu Symp Proc · 2024 · PMID 41726510
Social drivers of health significantly influence diabetes and hypertension outcomes. By taking into account patients' social and economic circumstances, healthcare systems can enhance both the quality and efficiency of c...
Social drivers of health significantly influence diabetes and hypertension outcomes. By taking into account patients' social and economic circumstances, healthcare systems can enhance both the quality and efficiency of care delivery, leading to improved health outcomes. This study aims to assess the concordance between patient-level social drivers data gathered from a patient-reported, health-related social needs survey and the data documented in electronic health records. A comparative analysis was conducted among 165 adults diagnosed with coexisting hypertension and uncontrolled diabetes from a singular academic health system. Each participant engaged in a standardized assessment of health-related social needs survey, and the corresponding electronic health record-based social drivers of health data were extracted. Concordance at the patient level for social drivers of health was assessed using Cohen's Kappa and percent agreement. Overall, agreement between the patient-reported social needs survey and electronic health records data was low, indicating only slight alignment across various social drivers of health domains. These findings suggest that relying solely on electronic health records data may underestimate the true prevalence of patient-reported social needs in this high-risk cohort with diabetes and hypertension. To ensure high-quality care delivery, there is a critical need for healthcare systems to develop more effective and sustainable methods for capturing social drivers of health data.
Racial and Ethnic Disparities in Identification of Cyanosis in ICU Settings.
Chaiken I, Sathe NA, Wurfel MM … +1 more , Wang LL
AMIA Annu Symp Proc · 2024 · PMID 41726509
Cyanosis is a discoloration of the skin arising from deoxygenated hemoglobin in the blood, caused by heart, lung, and blood diseases and treated with interventions including supplemental oxygen therapy. Cyanosis presents...
Cyanosis is a discoloration of the skin arising from deoxygenated hemoglobin in the blood, caused by heart, lung, and blood diseases and treated with interventions including supplemental oxygen therapy. Cyanosis presents as a bluish discoloration in light-skinned patients, but as a gray or white discoloration in dark-skinned patients. While prior work hints at the under-identification of cyanosis for people with black and brown skin, in this study, we quantify differences in cyanosis identification rates and associated clinical treatments by race/ethnicity. Leveraging EHR datasets from two hospital systems, we extract cyanosis mentions from clinical notes and compare cyanosis documentation rates by documented race/ethnicity. Cyanosis documentation was significantly less frequent for Black patients than White patients after adjusting for confounders. We measure impacts of cyanosis identification on provision of oxygen, vasopressors, and fluids. Adjusting for severity of a patient's condition, documentation of cyanosis was associated with faster provision of oxygen.
STop Clock for Automated Tracking (STAT) during Time-Critical Medical Work: Evaluating the Accuracy and Usability of an AI-Driven Automated Stop Clock.
Zellner KA, Yuan S, Ernst ER … +6 more , Arkowitz DW, Mun AH, Kim MS, Marsic I, Burd RS, Sarcevic A
AMIA Annu Symp Proc · 2024 · PMID 41726508
Delays and process inefficiencies during trauma resuscitation can contribute to adverse patient outcomes. While tracking elapsed time may improve the trauma team's temporal awareness and reduce delays, reliance on manual...
Delays and process inefficiencies during trauma resuscitation can contribute to adverse patient outcomes. While tracking elapsed time may improve the trauma team's temporal awareness and reduce delays, reliance on manual activation of stop clocks can introduce variability. To address this limitation, we implemented a computer vision-powered automatic stop clock designed to activate upon patient arrival without requiring manual input. We conducted a retrospective video review of 50 trauma resuscitations to assess how the clock was used in practice, followed by semi-structured interviews with nine trauma team members to elicit their feedback and perceptions. This study contributes to the broader discussion on AI-assisted clinical tools, highlighting the role of automation in supporting trauma teams, reducing variability in time tracking, and improving process efficiency.
Examining the Influence of Nurse Staffing on Delayed Insulin Administration in Acute Care Settings Using EHR and BCMA Data.
Rachman M, Yen PY
AMIA Annu Symp Proc · 2024 · PMID 41726507
Delayed insulin administration can lead to poor glycemic outcomes in patients with diabetes. Using EHR and BCMA data, we examined insulin administration patterns across different shifts and types of insulin, and the asso...
Delayed insulin administration can lead to poor glycemic outcomes in patients with diabetes. Using EHR and BCMA data, we examined insulin administration patterns across different shifts and types of insulin, and the association between nurse staffing and delayed administration. We analyzed a total of 650 subcutaneous insulin administration events from 96 patients. We found that 42.0% (n=397) of the insulins had delayed administration during 7a-3p shift. Long-acting insulin (Lantus) (64.6%) had more delays than other types of insulin, suggesting that the pharmacokinetics properties of these insulins may influence how nurses prioritized their insulin administrations. We also found that higher patient-to-nurse ratio was associated with delayed insulin administration; however, we did not find nursing skill mix was associated with delays. Lastly, we found patients with delayed insulin administration had poorer glycemic control. Our study demonstrates the need for evidence-based staffing that enables nurses to deliver timely insulin administration during high-demands periods.
Analyzing and Mitigating Model Drift in Acute Kidney Injury Prediction for Hospitalized Patients.
Xu Z, Li D, Xu Q … +3 more , Chan HY, Yu ASL, Liu M
AMIA Annu Symp Proc · 2024 · PMID 41726506
Artificial intelligence and machine learning are transforming healthcare by improving clinical risk predictions and diagnostic precision. However, their performance can be compromised by data drifts due to changes in pat...
Artificial intelligence and machine learning are transforming healthcare by improving clinical risk predictions and diagnostic precision. However, their performance can be compromised by data drifts due to changes in patient populations and evolving clinical practices. This study investigated performance drift in models predicting Acute Kidney Injury (AKI) using electronic health records from 249,749 inpatient encounters over ten years, analyzing performance across both the overall population and nine subgroups with unique health profiles. To mitigate the performance drift, we implemented two model updating strategies: an Overall Population Update (OPU) and a Specific Subgroup Update (SSU). Our results demonstrated significant reductions in drift, with OPU increasing the average area-under-the-precision-recall-curve (AUPRC) by 0.14 in the overall population and 0.11 across subgroups, and SSU improving the average AUPRC by 0.10 among subgroups. These findings highlight the importance of continuous model surveillance and adaptive updates to maintain reliable predictive performance in dynamic clinical environments.
Leveraging Provocative Design Methods to Address Implicit Bias in Clinical Interactions through Technology.
Mohanraj D, Langevin R, Shah L … +5 more , Sabin J, Wood BR, Pratt W, Weibel N, Hartzler AL
AMIA Annu Symp Proc · 2024 · PMID 41726505
Implicit bias impacts the quality of patient-clinician interactions, influencing patient outcomes and trust in healthcare. Most interventions to mitigate bias rely solely on expensive human assessments, rather than lever...
Implicit bias impacts the quality of patient-clinician interactions, influencing patient outcomes and trust in healthcare. Most interventions to mitigate bias rely solely on expensive human assessments, rather than leveraging AI technology with clinician input. To explore clinician-envisioned interventions, we conducted interviews with 16 primary care clinicians using provocative design methods to facilitate innovative ideation on using technology to address implicit bias. Themes from interviews included: patient communication monitoring, clinician self-awareness, systemic solutions, optimizing workflow, clinician education, and patient feedback. These envisioned interventions provide design considerations for technology-based implicit bias feedback tools. The broad range of innovative solutions generated by clinicians at various career stages reflects the utility of provocative design methods in unlocking creative thinking among a population that is not often encouraged to think beyond structured real-world constraints.
Identifying Missing IS-A Relations in SNOMED CT with Fine-Tuned Pre-trained Language Models and Non-lattice Subgraphs.
Hao X, Abeysinghe R, Shi J … +2 more , Zhang GQ, Cui L
AMIA Annu Symp Proc · 2024 · PMID 41726504
Ensuring the completeness of IS-A relations in SNOMED CT is crucial for maintaining its accuracy in clinical applications. In this study, we propose a hybrid approach leveraging non-lattice subgraphs and pre-trained lang...
Ensuring the completeness of IS-A relations in SNOMED CT is crucial for maintaining its accuracy in clinical applications. In this study, we propose a hybrid approach leveraging non-lattice subgraphs and pre-trained language models (PLMs) to identify missing IS-A relations in SNOMED CT. We fine-tuned four BERT-based models: BERT, DistillBERT, DeBERTa, and BioClinicalBERT, and four generative large language models (LLMs): BioMistral, Llama3, Gemma2, and Phi-4. Missing IS-A relations were identified through consensus predictions by all eight models. De-BERTa achieved the best performance (precision: 0.96, recall: 0.97, F1-score: 0.965) for IS-A relation prediction. Our approach identified 678 potential missing IS-A relations in SNOMED CT (March 2023 US Edition), of which 100 randomly selected cases were manually reviewed by a domain expert, confirming 93 as valid (93% precision). These results demonstrate the effectiveness of fine-tuned PLMs in detecting missing IS-A relations within non-lattice subgraphs, offering a promising avenue for improving SNOMED CT's quality.
About
- Frequency
- Sun
- Papers found
- 200
- RSS feed
- Subscribe