Benchmarking Modern Named Entity Recognition Techniques for Free-text Health Record De-identification

Electronic Health Records (EHRs) have become the primary form of medical data-keeping across the United States. Federal law restricts the sharing of any EHR data that contains protected health information (PHI). De-identification, the process of identifying and removing all PHI, is crucial for making EHR data publicly available for scientific research. This project explores several deep learning-based named entity recognition (NER) methods to determine which method(s) perform better on the de-identification task. We trained and tested our models on the i2b2 training dataset, and qualitatively assessed their performance using EHR data collected from a local hospital. We found that 1) BiLSTM-CRF represents the best-performing encoder/decoder combination, 2) character-embeddings and CRFs tend to improve precision at the price of recall, and 3) transformers alone under-perform as context encoders. Future work focused on structuring medical text may improve the extraction of semantic and syntactic information for the purposes of EHR de-identification.

TripClick: The Log Files of a Large Health Web Search Engine

Click logs are valuable resources for a variety of information retrieval (IR) tasks. This includes query understanding/analysis, as well as learning effective IR models particularly when the models require large amounts of training data. We release a large-scale domain-specific dataset of click logs, obtained from user interactions of the Trip Database health web search engine. Our click log dataset comprises approximately 5.2 million user interactions collected between 2013 and 2020. We use this dataset to create a standard IR evaluation benchmark -- TripClick -- with around 700,000 unique free-text queries and 1.3 million pairs of query-document relevance signals, whose relevance is estimated by two click-through models. As such, the collection is one of the few datasets offering the necessary data richness and scale to train neural IR models with a large amount of parameters, and notably the first in the health domain. Using TripClick, we conduct experiments to evaluate a variety of IR models, showing the benefits of exploiting this data to train neural architectures. In particular, the evaluation results show that the best performing neural IR model significantly improves the performance by a large margin relative to classical IR models, especially for more frequent queries.

Are "Undocumented Workers" the Same as "Illegal Aliens"? Disentangling Denotation and Connotation in Vector Spaces

In politics, neologisms are frequently invented for partisan objectives. For example, "undocumented workers" and "illegal aliens" refer to the same group of people (i.e., they have the same denotation), but they carry clearly different connotations. Examples like these have traditionally posed a challenge to reference-based semantic theories and led to increasing acceptance of alternative theories (e.g., Two-Factor Semantics) among philosophers and cognitive scientists. In NLP, however, popular pretrained models encode both denotation and connotation as one entangled representation. In this study, we propose an adversarial neural network that decomposes a pretrained representation as independent denotation and connotation representations. For intrinsic interpretability, we show that words with the same denotation but different connotations (e.g., "immigrants" vs. "aliens", "estate tax" vs. "death tax") move closer to each other in denotation space while moving further apart in connotation space. For extrinsic application, we train an information retrieval system with our disentangled representations and show that the denotation vectors improve the viewpoint diversity of document rankings.

Extracting Angina Symptoms from Clinical Notes Using Pre-Trained Transformer Architectures

Anginal symptoms can connote increased cardiac risk and a need for change in cardiovascular management. This study evaluated the potential to extract these symptoms from physician notes using the Bidirectional Encoder from Transformers language model fine-tuned on a domain-specific corpus. The history of present illness section of 459 expert annotated primary care physician notes from consecutive patients referred for cardiac testing without known atherosclerotic cardiovascular disease were included. Notes were annotated for positive and negative mentions of chest pain and shortness of breath characterization. The results demonstrate high sensitivity and specificity for the detection of chest pain or discomfort, substernal chest pain, shortness of breath, and dyspnea on exertion. Small sample size limited extracting factors related to provocation and palliation of chest pain. This study provides a promising starting point for the natural language processing of physician notes to characterize clinically actionable anginal symptoms.

A Transformer-based Framework for Multivariate Time Series Representation Learning

In this work we propose for the first time a transformer-based framework for unsupervised representation learning of multivariate time series. Pre-trained models can be potentially used for downstream tasks such as regression and classification, forecasting and missing value imputation. By evaluating our models on several benchmark datasets for multivariate time series regression and classification, we show that not only does our modeling approach represent the most successful method employing unsupervised learning of multivariate time series presented to date, but also that it exceeds the current state-of-the-art performance of supervised methods; it does so even when the number of training samples is very limited, while offering computational efficiency. Finally, we demonstrate that unsupervised pre-training of our transformer models offers a substantial performance benefit over fully supervised learning, even without leveraging additional unlabeled data, i.e., by reusing the same data samples through the unsupervised objective.

Brown University at TREC Deep Learning 2019

This paper describes Brown University's submission to the TREC 2019 Deep Learning track. We followed a 2-phase method for producing a ranking of passages for a given input query: In the the first phase, the user's query is expanded by appending 3 queries generated by a transformer model which was trained to rephrase an input query into semantically similar queries. The expanded query can exhibit greater similarity in surface form and vocabulary overlap with the passages of interest and can therefore serve as enriched input to any downstream information retrieval method. In the second phase, we use a BERT-based model pre-trained for language modeling but fine-tuned for query - document relevance prediction to compute relevance scores for a set of 1000 candidate passages per query and subsequently obtain a ranking of passages by sorting them based on the predicted relevance scores. According to the results published in the official Overview of the TREC Deep Learning Track 2019, our team ranked 3rd in the passage retrieval task (including full ranking and re-ranking), and 2nd when considering only re-ranking submissions.

Diagnosis Prevalence vs. Efficacy in Machine-learning Based Diagnostic Decision Support

Many recent studies use machine learning to predict a small number of ICD-9-CM codes. In practice, on the other hand, physicians have to consider a broader range of diagnoses. This study aims to put these previously incongruent evaluation settings on a more equal footing by predicting ICD-9-CM codes based on electronic health record properties and demonstrating the relationship between diagnosis prevalence and system performance. We extracted patient features from the MIMIC-III dataset for each admission. We trained and evaluated 43 different machine learning classifiers. Among this pool, the most successful classifier was a Multi-Layer Perceptron. In accordance with general machine learning expectation, we observed all classifiers' F1 scores to drop as disease prevalence decreased. Scores fell from 0.28 for the 50 most prevalent ICD-9-CM codes to 0.03 for the 1000 most prevalent ICD-9-CM codes. Statistical analyses showed a moderate positive correlation between disease prevalence and efficacy (0.5866).

Wasserstein Adversarial Autoencoders for Knowledge Graph Embedding based Drug-Drug Interaction Prediction

Interaction between pharmacological agents can trigger unexpected adverse events. Capturing richer and more comprehensive information about drug-drug interactions (DDI) is one of the key tasks in public health and drug development. Recently, several knowledge graph embedding approaches have received increasing attention in the DDI domain due to their capability of projecting drugs and interactions into a low-dimensional feature space for predicting links and classifying triplets. However, existing methods only apply a uniformly random mode to construct negative samples. As a consequence, these samples are often too simplistic to train an effective model. In this paper, we propose a new knowledge graph embedding framework by introducing adversarial autoencoders (AAE) based on Wasserstein distances and Gumbel-Softmax relaxation for drug-drug interactions tasks. In our framework, the autoencoder is employed to generate high-quality negative samples and the hidden vector of the autoencoder is regarded as a plausible drug candidate. Afterwards, the discriminator learns the embeddings of drugs and interactions based on both positive and negative triplets. Meanwhile, in order to solve vanishing gradient problems on the discrete representation--an inherent flaw in traditional generative models--we utilize the Gumbel-Softmax relaxation and the Wasserstein distance to train the embedding model steadily. We empirically evaluate our method on two tasks, link prediction and DDI classification. The experimental results show that our framework can attain significant improvements and noticeably outperform competitive baselines.

DC3 -- A Diagnostic Case Challenge Collection for Clinical Decision Support

In clinical care, obtaining a correct diagnosis is the first step towards successful treatment and, ultimately, recovery. Depending on the complexity of the case, the diagnostic phase can be lengthy and ridden with errors and delays. Such errors have a high likelihood to cause patients severe harm or even lead to their death and are estimated to cost the U.S. healthcare system several hundred billion dollars each year. To avoid diagnostic errors, physicians increasingly rely on diagnostic decision support systems drawing from heuristics, historic cases, textbooks, clinical guidelines and scholarly biomedical literature. The evaluation of such systems, however, is often conducted in an ad-hoc fashion, using non-transparent methodology, and proprietary data. This paper presents DC3, a collection of 31 extremely difficult diagnostic case challenges, manually compiled and solved by clinical experts. For each case, we present a number of temporally ordered physician-generated observations alongside the eventually confirmed true diagnosis. We additionally provide inferred dense relevance judgments for these cases among the PubMed collection of 27 million scholarly biomedical articles.