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

Mobile personal health records (mPHR) are smartphone apps that grant patients portable and continuous access to their medical records, thereby increasing the potential for patients to play an active role in managing their health. An extensive body of literature has focused on understanding user experiences with web-based tethered PHRs (i.e., patient portals) offered by healthcare organizations. However, patients' opinions of smartphone-based PHRs have received less attention. To address this gap, we used a computationally-guided qualitative analysis approach to analyze user reviews of six tethered and four interconnected mPHR apps available on both Google Play and Apple app stores. This approach resulted in identifying dimensions of user experiences related to usability, usefulness, and important features to users. Our findings reveal many similarities in user experiences for HCO-tethered and HCO-independent interconnected PHRs. However, there are some differences in user experiences between the types of PHRs and the different devices and platforms.

Screening mammogram is a standard and cost-efficient imaging procedure to measure breast cancer risk among 45+ year old women. Quantifying breast arterial calcification (BAC) from screening mammograms is a non-invasive and cost-efficient approach to assess the future risk of adverse cardiovascular events among women, such as heart attack and stroke. However, segmentation of breast arterial calcification is an involved task and poses several technical challenges such as extremely small BAC finding, low breast arteries to breast area ratio in the mammogram images, tissue features such as breast folds and heterogeneous density, have very similar imaging appearance. In this work, we aim to address the shortcomings of existing SOTA methods, e.g., SCUNet, and analyze the comparative performance. Given the fact that we will not be able to simply resize mammogram to preserve the microscopic BAC details, we adopted a patch-based methodology for segmentation using the original resolution which may hinder the model understanding of whole mammogram. We propose a multi-task learning approach for patch-based BAC segmentation by adding an auxiliary task of patch position prediction which forces the model to learn breast anatomy to comprehend the locations where BAC will not occur, such as breast boundary. The proposed method achieves state-of-the-art performance compared to the baselines. To demonstrate the utility, we also validate our method on external data and provide survival analysis for adverse cardiac events based on difference in BAC score and provide a comparison with coronary calcium score (CAC).

The delivery of effective targeted therapies requires comprehensive analyses of the molecular profiling of tumors and matching with clinical phenotypes in the context of existing knowledge described in biomedical literature, registries, and knowledge bases. We evaluated the performance of natural language processing (NLP) approaches in supporting knowledge retrieval and synthesis from the biomedical literature. We tested PubTator 3.0, Bidirectional Encoder Representations from Transformers (BERT), and Large Language Models (LLMs) and evaluated their ability to support named entity recognition (NER) and relation extraction (RE) from biomedical texts. PubTator 3.0 and the BioBERT model performed best in the NER task (best F1-score 0.93 and 0.89, respectively), while BioBERT outperformed all other solutions in the RE task (best F1-score 0.79) and a specific use case it was applied to by recognizing nearly all entity mentions and most of the relations. Our findings support the use of AI-assisted approaches in facilitating precision oncology decision-making.

Electronic health records (EHRs) are information systems designed to collect and manage clinical data in order to support various clinical activities. They have emerged as valuable sources of data for outcomes research, offering vast repositories of patient information for analysis. Definitions for pediatric sepsis diagnosis are ambiguous, resulting in delayed diagnosis and treatment, highlighting the need for precise and efficient patient categorizing techniques. Nevertheless, the use of EHRs in research poses challenges. Although EHRs were originally created to document patient encounters, the medical coding was designed to satisfy billing requirements. As a result, EHR data may lack granularity, potentially leading to misclassification and incomplete representation of patient conditions. We compared data-driven ICD code categories to chart review using probabilistic graphical models (PGMs) due to their ability to handle uncertainty and incorporate prior knowledge. Overall, this paper demonstrates the potential of using PGMs to address these challenges and improve the analysis of ICD codes for sepsis outcomes research.

In aims to uncover insights into medical decision-making embedded within observational data from clinical settings, we present a novel application of Inverse Reinforcement Learning (IRL) that identifies suboptimal clinician actions based on the actions of their peers. This approach centers two stages of IRL with an intermediate step to prune trajectories displaying behavior that deviates significantly from the consensus. This enables us to effectively identify clinical priorities and values from ICU data containing both optimal and suboptimal clinician decisions. We observe that the benefits of removing suboptimal actions vary by disease and differentially impact certain demographic groups.

Portable genomic sequencers such as Oxford Nanopore's MinION enable real-time applications in clinical and environmental health. However, there is a bottleneck in the downstream analytics when bioinformatics pipelines are unavailable, e.g., when cloud processing is unreachable due to absence of Internet connection, or only low-end computing devices can be carried on site. Here we present a platform-friendly software for portable metagenomic analysis of Nanopore data, the Oligomer-based Classifier of Taxonomic Operational and Pan-genome Units via Singletons (OCTOPUS). OCTOPUS is written in Java, reimplements several features of the popular Kraken2 and KrakenUniq software, with original components for improving metagenomics classification on incomplete/sampled reference databases, making it ideal for running on smartphones or tablets. OCTOPUS obtains sensitivity and precision comparable to Kraken2, while dramatically decreasing (4- to 16-fold) the false positive rate, and yielding high correlation on real-word data. OCTOPUS is available along with customized databases at https://github.com/DataIntellSystLab/OCTOPUS and https://github.com/Ruiz-HCI-Lab/OctopusMobile.

Large electronic health records (EHR) have been widely implemented and are available for research activities. The magnitude of such databases often requires storage and computing infrastructure that are distributed at different sites. Restrictions on data-sharing due to privacy concerns have been another driving force behind the development of a large class of distributed and/or federated machine learning methods. While missing data problem is also present in distributed EHRs, albeit potentially more complex, distributed multiple imputation (MI) methods have not received as much attention. An important advantage of distributed MI, as well as distributed analysis, is that it allows researchers to borrow information across data sites, mitigating potential fairness issues for minority groups that do not have enough volume at certain sites. In this paper, we propose a communication-efficient and privacy-preserving distributed MI algorithms for variables that are missing not at random.

Collecting images of rare dermatological diseases for machine learning detection applications is a costly, laborious task. It is difficult to collect enough images of these diagnoses to avoid the risk of low accuracy "in the wild". One of the sources of bias in these networks is irrelevant background pixel data. These pixels necessarily have no clinical significance, yet Deep Neural Networks will make weak correlations based on that information. To reduce their ability to do this, we introduce a masking augmentation algorithm, InfoMax-Cutout. It employs unsupervised Information Maximization losses to mask out background pixels. InfoMax-Cutout increased accuracy on classifying 319 diagnoses by 0.76%. These features generalized to an unseen diagnosis task (Fitzpatrick 17k), improving accuracy over a baseline by 43.3% and reducing Gini inequality by 20.9%. This approach of learning to separate out background pixels can increase accuracy in detecting diseases in Lower and Middle Income Countries.

Electronic health records (EHRs) contain valuable patient data for health-related prediction tasks, such as disease prediction. Traditional approaches rely on supervised learning methods that require large labeled datasets, which can be expensive and challenging to obtain. In this study, we investigate the feasibility of applying Large Language Models (LLMs) to convert structured patient visit data (e.g., diagnoses, labs, prescriptions) into natural language narratives. We evaluate the zero-shot and few-shot performance of LLMs using various EHR-prediction-oriented prompting strategies. Furthermore, we propose a novel approach that utilizes LLM agents with different roles: a predictor agent that makes predictions and generates reasoning processes and a critic agent that analyzes incorrect predictions and provides guidance for improving the reasoning of the predictor agent. Our results demonstrate that with the proposed approach, LLMs can achieve decent few-shot performance compared to traditional supervised learning methods in EHR-based disease predictions, suggesting its potential for health-oriented applications.

Social determinants of health (SDoH) significantly impacts health outcomes and contributes to perpetuating health disparities across healthcare applications. However, automatic extraction of SDoH information from Electronic Health Records (EHRs) is challenging due to the unstructured nature of clinical narratives that contain SDoH related information. Recent advances in Large Language Models (LLMs) have shown great promise for automated SDoH extraction. However, their performance suffers for the imbalanced SDoH categories due to the data scarcity issues. To address this, we propose an innovative approach that augments LLMs with semantic knowledge obtained from the Unified Medical Language Systems (UMLS). This strategy enriches the feature representations of imbalanced SDoH classes, leading to accurate SDoH extraction. More specifically, our proposed data augmentation strategy generates semantically enriched clinical narratives at the LLM pre-finetuning stage. This approach enables the LLM to better adapt to the target data and leads to a good initialization for the finetuning stage. Through extensive experiments using publicly available MIMIC-SDoH data, the proposed approach demonstrates significant improvement in results for the SDoH extraction, especially for the imbalanced classes.