Treatment initiation prediction by EHR mapped PPD tensor based convolutional neural networks boosting algorithm.

Electronic health records contain patient's information that can be used for health analytics tasks such as disease detection, disease progression prediction, patient profiling, etc. Traditional machine learning or deep learning methods treat EHR entities as individual features, and no relationships between them are taken into consideration. We propose to evaluate the relationships between EHR features and map them into Procedures, Prescriptions, and Diagnoses (PPD) tensor data, which can be formatted as images. The mapped images are then fed into deep convolutional networks for local pattern and feature learning. We add this relationship-learning part as a boosting module on a commonly used classical machine learning model. Experiments were performed on a Chronic Lymphocytic Leukemia dataset for treatment initiation prediction. Experimental results show that the proposed approach has better real world modeling performance than the baseline models in terms of prediction precision.

Exploiting Causality for Improved Prediction of Patient Volumes by Gaussian Processes.

Estimating and surveillance volumes of patients are of great importance for public health and resource allocation. In many situations, the change of these volumes is correlated with many factors, e.g., seasonal environmental variables, medicine sales, and patient medical claims. It is often of interest to predict patient volumes and to that end, discovering causalities can improve the prediction accuracy. Correlations do not imply causations and they can be spurious, which in turn may entail deterioration of prediction performance if the prediction is based on them. By contrast, in this paper, we propose an approach for prediction based on causalities discovered by Gaussian processes. Our interest is in estimating volumes of patients that suffer from allergy and where the model and the results are highly interpretable. In selecting features, instead of only using correlation, we take causal information into account. Specifically, we adopt the Gaussian processes-based convergent cross mapping framework for causal discovery which is proven to be more reliable than the Granger causality when time series are coupled. Moreover, we introduce a novel method for selecting the history or look-back length of features from the perspective of a dynamical system in a principled manner. The quasi-periodicities that commonly exist in observations of volumes of patients and environment variables can readily be accommodated. Further, the proposed method performs well even in cases when the data are scarce. Also, the approach can be modified without much difficulty to forecast other types of patient volumes. We validate the method with synthetic and real-world datasets.

Leveraging graph-based hierarchical medical entity embedding for healthcare applications.

Automatic representation learning of key entities in electronic health record (EHR) data is a critical step for healthcare data mining that turns heterogeneous medical records into structured and actionable information. Here we propose ME2Vec, an algorithmic framework for learning continuous low-dimensional embedding vectors of the most common entities in EHR: medical services, doctors, and patients. ME2Vec features a hierarchical structure that encapsulates different node embedding schemes to cater for the unique characteristic of each medical entity. To embed medical services, we employ a biased-random-walk-based node embedding that leverages the irregular time intervals of medical services in EHR to embody their relative importance. To embed doctors and patients, we adhere to the principle "it's what you do that defines you" and derive their embeddings based on their interactions with other types of entities through graph neural network and proximity-preserving network embedding, respectively. Using real-world clinical data, we demonstrate the efficacy of ME2Vec over competitive baselines on diagnosis prediction, readmission prediction, as well as recommending doctors to patients based on their medical conditions. In addition, medical service embeddings pretrained using ME2Vec can substantially improve the performance of sequential models in predicting patients clinical outcomes. Overall, ME2Vec can serve as a general-purpose representation learning algorithm for EHR data and benefit various downstream tasks in terms of both performance and interpretability.

On exchangeable multinomial distributions.

We derive an expression for the joint distribution of exchangeable multinomial random variables, which generalizes the multinomial distribution based on independent trials while retaining some of its important properties. Unlike de Finneti's representation theorem for a binary sequence, the exchangeable multinomial distribution derived here does not require that the finite set of random variables under consideration be a subset of an infinite sequence. Using expressions for higher moments and correlations, we show that the covariance matrix for exchangeable multinomial data has a different form from that usually assumed in the literature, and we analyse data from developmental toxicology studies. The proposed analyses have been implemented in R and are available on CRAN in the CorrBin package.