AMIA ... Annual Symposium proceedings. AMIA Symposium最新文献

Chronic gastrointestinal (GI) conditions, such as inflammatory bowel diseases (IBD), offer a promising opportunity to create classification systems that can enhance the accuracy of predicting the most effective therapies and prognosis for each patient. Here, we present a novel methodology to explore disease subtypes using our open-sourced BiomedSciAI toolkit. Applying methods available in this toolkit on the UK Biobank, including subpopulation-based feature selection and multi-dimensional subset scanning, we aimed to discover unique subgroups from GI surgery cohorts. Of a 12,073-patient cohort, a subgroup of 440 IBD patients was discovered with an increased risk of a subsequent GI surgery (OR: 2.21, 95% CI [1.81-2.69]). We iteratively demonstrate the discovery process using an additional cohort (with a narrower definition of GI surgery). Our results show that the iterative process can refine the subgroup discovery process and generate novel hypotheses to investigate determinants of treatment response.

Objective: To implement an open source, free, and easily deployable high throughput natural language processing module to extract concepts from clinician notes and map them to Fast Healthcare Interoperability Resources (FHIR). Materials and Methods: Using a popular open-source NLP tool (Apache cTAKES), we create FHIR resources that use modifier extensions to represent negation and NLP sourcing, and another extension to represent provenance of extracted concepts. Results: The SMART Text2FHIR Pipeline is an open-source tool, released through standard package managers, and publicly available container images that implement the mappings, enabling ready conversion of clinical text to FHIR. Discussion: With the increased data liquidity because of new interoperability regulations, NLP processes that can output FHIR can enable a common language for transporting structured and unstructured data. This framework can be valuable for critical public health or clinical research use cases. Conclusion: Future work should include mapping more categories of NLP-extracted information into FHIR resources and mappings from additional open-source NLP tools.

Lymphoma is one of the most common types of cancer for children (ages 0 to 19). Due to the reduced radiation exposure, PET/MR systems that allow simultaneous PET and MR imaging have become the standard of care for diagnosing cancers and monitoring tumor response to therapy in the pediatric population. In this work, we developed a multimodal deep learning algorithm for automatic pediatric lymphoma detection using PET and MRI. Through innovative designs such as standardized uptake value (SUV) guided tumor candidate generation, location aware classification model learning and weighted multimodal feature fusion, our algorithm can be effectively trained with limited data and achieved superior tumor detection performance over the state-of-the-art in our experiments.

Individuals diagnosed with autism spectrum disorder (ASD) are at a higher risk for mental health concerns including suicidal thoughts and behaviors (STB). Limited studies have focused on suicidal risk factors that are more prevalent or unique to the population with ASD. This study sought to characterize and classify youth presenting to the psychiatric emergency department (ED) for a chief complaint of STB. The results of this study validated that a high number of patients with ASD present to the ED with STB. There were important differences in clinical characteristics to those with ASD versus those without. Clinical features that showed important impact in predicting high suicide risk in the ASD cases include elements of the mental status exam such as affect, trauma symptoms, abuse history, and auditory hallucinations. Focused attention is needed on these unique differences in ASD cases so that suicide risk level can be appropriately and promptly addressed.

Determining clinically relevant physiological states from multivariate time-series data with missing values is essential for providing appropriate treatment for acute conditions such as Traumatic Brain Injury (TBI), respiratory failure, and heart failure. Utilizing non-temporal clustering or data imputation and aggregation techniques may lead to loss of valuable information and biased analyses. In our study, we apply the SLAC-Time algorithm, an innovative self-supervision-based approach that maintains data integrity by avoiding imputation or aggregation, offering a more useful representation of acute patient states. By using SLAC-Time to cluster data in a large research dataset, we identified three distinct TBI physiological states and their specific feature profiles. We employed various clustering evaluation metrics and incorporated input from a clinical domain expert to validate and interpret the identified physiological states. Further, we discovered how specific clinical events and interventions can influence patient states and state transitions.

The proliferation of Deep Learning (DL)-based methods for radiographic image analysis has created a great demand for expert-labeled radiology data. Recent self-supervised frameworks have alleviated the need for expert labeling by obtaining supervision from associated radiology reports. These frameworks, however, struggle to distinguish the subtle differences between different pathologies in medical images. Additionally, many of them do not provide interpretation between image regions and text, making it difficult for radiologists to assess model predictions. In this work, we propose Local Region Contrastive Learning (LRCLR), a flexible fine-tuning framework that adds layers for significant image region selection as well as cross-modality interaction. Our results on an external validation set of chest x-rays suggest that LRCLR identifies significant local image regions and provides meaningful interpretation against radiology text while improving zero-shot performance on several chest x-ray medical findings.

Clinical trials are indispensable in developing new treatments, but they face obstacles in patient recruitment and retention, hindering the enrollment of necessary participants. To tackle these challenges, deep learning frameworks have been created to match patients to trials. These frameworks calculate the similarity between patients and clinical trial eligibility criteria, considering the discrepancy between inclusion and exclusion criteria. Recent studies have shown that these frameworks outperform earlier approaches. However, deep learning models may raise fairness issues in patient-trial matching when certain sensitive groups of individuals are underrepresented in clinical trials, leading to incomplete or inaccurate data and potential harm. To tackle the issue of fairness, this work proposes a fair patient-trial matching framework by generating a patient-criterion level fairness constraint. The proposed framework considers the inconsistency between the embedding of inclusion and exclusion criteria among patients of different sensitive groups. The experimental results on real-world patient-trial and patient-criterion matching tasks demonstrate that the proposed framework can successfully alleviate the predictions that tend to be biased.

Sexual gender minorities, including lesbian, gay, and bisexual (LGB) individuals face unique challenges due to discrimination, stigma, and marginalization, which negatively impact their well-being. Electronic health record (EHR) systems present an opportunity for LGB research, but accurately identifying LGB individuals in EHRs is challenging. Our study developed and validated a rule-based computable phenotype (CP) to identify LGB individuals and their subgroups using both structured data and unstructured clinical narratives from a large integrated health system. Validating against a sample of 537 chart-reviewed patients, our three best performing CP algorithms balancing different performance metrics, each achieved sensitivity of 1.000, PPV of 0.982, and F1-score of 0.875 in identifying LGB individuals, respectively. Applying the three best-performing CPs, our study also found that the LGB population is younger and experiences a disproportionate burden of adverse health outcomes, particularly mental health distress.

Workflow fragmentation, defined as task switching, may be one proxy to quantify electronic health record (EHR) documentation burden in the emergency department (ED). Few measures have been operationalized to evaluate task switching at scale. Theoretically grounded in the time-based resource-sharing model (TBRSM) which conceives task switching as proportional to the cognitive load experienced, we describe the functional relationship between cognitive load and the time and effort constructs previously applied for measuring documentation burden. We present a computational framework, COMBINE, to evaluate multilevel task switching in the ED using EHR event logs. Based on this framework, we conducted a descriptive analysis on task switching among 63 full-time ED physicians from one ED site using EHR event logs extracted between April-June 2021 (n=2,068,605 events) which were matched to scheduled shifts (n=952). On average, we found a high volume of event-level (185.8±75.3/hr) and within-(6.6±1.7/chart) and between-patient chart (27.5±23.6/hr) switching per shift worked.

Computerized provider order entry (CPOE) systems have been cited as a significant contributor to clinician burden. Vendor-derived measures and data sets have been developed to help with optimization of CPOE systems. We describe how we analyzed vendor-derived Order Friction (OF) EHR log data at our health system and propose a practical approach for optimizing CPOE systems by reducing OF. We also conducted a pre-post intervention study using OF data to evaluate the impact of defaulting the frequency of urine, stool and nasal swab tests and found that all modified orders had significantly fewer changes required per order (p<0.01). Our proposed approach is a six-step process: 1) understand the ordering process, 2) understand OF data elements contextually, 3) explore ordering user-level factors, 4) evaluate order volume and friction from different order sources, 5) optimize order-level design, 6) identify high volume alerts to evaluate for appropriateness.