Dataset Information

Subgraph augmented non-negative tensor factorization (SANTF) for modeling clinical narrative text.

ABSTRACT:

Objective

Extracting medical knowledge from electronic medical records requires automated approaches to combat scalability limitations and selection biases. However, existing machine learning approaches are often regarded by clinicians as black boxes. Moreover, training data for these automated approaches at often sparsely annotated at best. The authors target unsupervised learning for modeling clinical narrative text, aiming at improving both accuracy and interpretability.

Methods

The authors introduce a novel framework named subgraph augmented non-negative tensor factorization (SANTF). In addition to relying on atomic features (e.g., words in clinical narrative text), SANTF automatically mines higher-order features (e.g., relations of lymphoid cells expressing antigens) from clinical narrative text by converting sentences into a graph representation and identifying important subgraphs. The authors compose a tensor using patients, higher-order features, and atomic features as its respective modes. We then apply non-negative tensor factorization to cluster patients, and simultaneously identify latent groups of higher-order features that link to patient clusters, as in clinical guidelines where a panel of immunophenotypic features and laboratory results are used to specify diagnostic criteria.

Results and conclusion

SANTF demonstrated over 10% improvement in averaged F-measure on patient clustering compared to widely used non-negative matrix factorization (NMF) and k-means clustering methods. Multiple baselines were established by modeling patient data using patient-by-features matrices with different feature configurations and then performing NMF or k-means to cluster patients. Feature analysis identified latent groups of higher-order features that lead to medical insights. We also found that the latent groups of atomic features help to better correlate the latent groups of higher-order features.

SUBMITTER: Luo Y

PROVIDER: S-EPMC4986663 | biostudies-literature | 2015 Sep

REPOSITORIES: biostudies-literature

ACCESS DATA

Publications

Subgraph augmented non-negative tensor factorization (SANTF) for modeling clinical narrative text.

Luo Yuan Y Xin Yu Y Hochberg Ephraim E Joshi Rohit R Uzuner Ozlem O Szolovits Peter P

Journal of the American Medical Informatics Association : JAMIA 20150409 5

<h4>Objective</h4>Extracting medical knowledge from electronic medical records requires automated approaches to combat scalability limitations and selection biases. However, existing machine learning approaches are often regarded by clinicians as black boxes. Moreover, training data for these automated approaches at often sparsely annotated at best. The authors target unsupervised learning for modeling clinical narrative text, aiming at improving both accuracy and interpretability.<h4>Methods</h ...[more]

PMID: 25862765

Similar Datasets

Project description:Background:Matrix factorization is a well established pattern discovery tool that has seen numerous applications in biomedical data analytics, such as gene expression co-clustering, patient stratification, and gene-disease association mining. Matrix factorization learns a latent data model that takes a data matrix and transforms it into a latent feature space enabling generalization, noise removal and feature discovery. However, factorization algorithms are numerically intensive, and hence there is a pressing challenge to scale current algorithms to work with large datasets. Our focus in this paper is matrix tri-factorization, a popular method that is not limited by the assumption of standard matrix factorization about data residing in one latent space. Matrix tri-factorization solves this by inferring a separate latent space for each dimension in a data matrix, and a latent mapping of interactions between the inferred spaces, making the approach particularly suitable for biomedical data mining. Results:We developed a block-wise approach for latent factor learning in matrix tri-factorization. The approach partitions a data matrix into disjoint submatrices that are treated independently and fed into a parallel factorization system. An appealing property of the proposed approach is its mathematical equivalence with serial matrix tri-factorization. In a study on large biomedical datasets we show that our approach scales well on multi-processor and multi-GPU architectures. On a four-GPU system we demonstrate that our approach can be more than 100-times faster than its single-processor counterpart. Conclusions:A general approach for scaling non-negative matrix tri-factorization is proposed. The approach is especially useful parallel matrix factorization implemented in a multi-GPU environment. We expect the new approach will be useful in emerging procedures for latent factor analysis, notably for data integration, where many large data matrices need to be collectively factorized.

Project description:ObjectiveTo demonstrate how non-negative matrix factorization can be used to learn a temporal topic model over a large collection of primary care clinical notes, characterizing diverse COVID-19 pandemic effects on the physical/mental/social health of residents of Toronto, Canada.Materials and methodsThe study employs a retrospective open cohort design, consisting of 382,666 primary care progress notes from 44,828 patients, 54 physicians, and 12 clinics collected 01/01/2017 through 31/12/2020. Non-negative matrix factorization uncovers a meaningful latent topical structure permeating the corpus of primary care notes. The learned latent topical basis is transformed into a multivariate time series data structure. Time series methods and plots showcase the evolution/dynamics of learned topics over the study period and allow the identification of COVID-19 pandemic effects. We perform several post-hoc checks of model robustness to increase trust that descriptive/unsupervised inferences are stable over hyper-parameter configurations and/or data perturbations.ResultsTemporal topic modelling uncovers a myriad of pandemic-related effects from the expressive clinical text data. In terms of direct effects on patient-health, topics encoding respiratory disease symptoms display altered dynamics during the pandemic year. Further, the pandemic was associated with a multitude of indirect patient-level effects on topical domains representing mental health, sleep, social and familial dynamics, measurement of vitals/labs, uptake of prevention/screening maneuvers, and referrals to medical specialists. Finally, topic models capture changes in primary care practice patterns resulting from the pandemic, including changes in EMR documentation strategies and the uptake of telemedicine.ConclusionTemporal topic modelling applied to a large corpus of rich primary care clinical text data, can identify a meaningful topical/thematic summarization which can provide policymakers and public health stakeholders a passive, cost-effective, technology for understanding holistic impacts of the COVID-19 pandemic on the primary healthcare system and community/public-health.

Dataset Information

Subgraph augmented non-negative tensor factorization (SANTF) for modeling clinical narrative text.

Objective

Methods

Results and conclusion

Publications

Subgraph augmented non-negative tensor factorization (SANTF) for modeling clinical narrative text.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets