Yuan Luo

Yikuan Li, Ramsey M. Wehbe, Faraz S. Ahmad et al.

Objective: Clinical knowledge enriched transformer models (e.g., ClinicalBERT) have state-of-the-art results on clinical NLP (natural language processing) tasks. One of the core limitations of these transformer models is the substantial memory consumption due to their full self-attention mechanism, which leads to the performance degradation in long clinical texts. To overcome this, we propose to leverage long-sequence transformer models (e.g., Longformer and BigBird), which extend the maximum input sequence length from 512 to 4096, to enhance the ability to model long-term dependencies in long clinical texts. Materials and Methods: Inspired by the success of long sequence transformer models and the fact that clinical notes are mostly long, we introduce two domain enriched language models, Clinical-Longformer and Clinical-BigBird, which are pre-trained on a large-scale clinical corpus. We evaluate both language models using 10 baseline tasks including named entity recognition, question answering, natural language inference, and document classification tasks. Results: The results demonstrate that Clinical-Longformer and Clinical-BigBird consistently and significantly outperform ClinicalBERT and other short-sequence transformers in all 10 downstream tasks and achieve new state-of-the-art results. Discussion: Our pre-trained language models provide the bedrock for clinical NLP using long texts. We have made our source code available at https://github.com/luoyuanlab/Clinical-Longformer, and the pre-trained models available for public download at: https://huggingface.co/yikuan8/Clinical-Longformer. Conclusion: This study demonstrates that clinical knowledge enriched long-sequence transformers are able to learn long-term dependencies in long clinical text. Our methods can also inspire the development of other domain-enriched long-sequence transformers.

Zheng Yu, Yikuan Li, Joseph Kim et al.

Dynamic diagnosis is desirable when medical tests are costly or time-consuming. In this work, we use reinforcement learning (RL) to find a dynamic policy that selects lab test panels sequentially based on previous observations, ensuring accurate testing at a low cost. Clinical diagnostic data are often highly imbalanced; therefore, we aim to maximize the $F_1$ score instead of the error rate. However, optimizing the non-concave $F_1$ score is not a classic RL problem, thus invalidates standard RL methods. To remedy this issue, we develop a reward shaping approach, leveraging properties of the $F_1$ score and duality of policy optimization, to provably find the set of all Pareto-optimal policies for budget-constrained $F_1$ score maximization. To handle the combinatorially complex state space, we propose a Semi-Model-based Deep Diagnosis Policy Optimization (SM-DDPO) framework that is compatible with end-to-end training and online learning. SM-DDPO is tested on diverse clinical tasks: ferritin abnormality detection, sepsis mortality prediction, and acute kidney injury diagnosis. Experiments with real-world data validate that SM-DDPO trains efficiently and identifies all Pareto-front solutions. Across all tasks, SM-DDPO is able to achieve state-of-the-art diagnosis accuracy (in some cases higher than conventional methods) with up to $85\%$ reduction in testing cost. The code is available at [https://github.com/Zheng321/Deep-Reinforcement-Learning-for-Cost-Effective-Medical-Diagnosis].

Yuan Luo

Hypergraph neural networks can model multi-way connections among nodes of the graphs, which are common in real-world applications such as genetic medicine. In particular, genetic pathways or gene sets encode molecular functions driven by multiple genes, naturally represented as hyperedges. Thus, hypergraph-guided embedding can capture functional relations in learned representations. Existing hypergraph neural network models often focus on node-level or graph-level inference. There is an unmet need in learning powerful representations of subgraphs of hypergraphs in real-world applications. For example, a cancer patient can be viewed as a subgraph of genes harboring mutations in the patient, while all the genes are connected by hyperedges that correspond to pathways representing specific molecular functions. For accurate inductive subgraph prediction, we propose SubHypergraph Inductive Neural nEtwork (SHINE). SHINE uses informative genetic pathways that encode molecular functions as hyperedges to connect genes as nodes. SHINE jointly optimizes the objectives of end-to-end subgraph classification and hypergraph nodes' similarity regularization. SHINE simultaneously learns representations for both genes and pathways using strongly dual attention message passing. The learned representations are aggregated via a subgraph attention layer and used to train a multilayer perceptron for inductive subgraph inferencing. We evaluated SHINE against a wide array of state-of-the-art (hyper)graph neural networks, XGBoost, NMF and polygenic risk score models, using large scale NGS and curated datasets. SHINE outperformed all comparison models significantly, and yielded interpretable disease models with functional insights.

Chengsheng Mao, Jie Xu, Luke Rasmussen et al.

Objective: We develop a deep learning framework based on the pre-trained Bidirectional Encoder Representations from Transformers (BERT) model using unstructured clinical notes from electronic health records (EHRs) to predict the risk of disease progression from Mild Cognitive Impairment (MCI) to Alzheimer's Disease (AD). Materials and Methods: We identified 3657 patients diagnosed with MCI together with their progress notes from Northwestern Medicine Enterprise Data Warehouse (NMEDW) between 2000-2020. The progress notes no later than the first MCI diagnosis were used for the prediction. We first preprocessed the notes by deidentification, cleaning and splitting, and then pretrained a BERT model for AD (AD-BERT) based on the publicly available Bio+Clinical BERT on the preprocessed notes. The embeddings of all the sections of a patient's notes processed by AD-BERT were combined by MaxPooling to compute the probability of MCI-to-AD progression. For replication, we conducted a similar set of experiments on 2563 MCI patients identified at Weill Cornell Medicine (WCM) during the same timeframe. Results: Compared with the 7 baseline models, the AD-BERT model achieved the best performance on both datasets, with Area Under receiver operating characteristic Curve (AUC) of 0.8170 and F1 score of 0.4178 on NMEDW dataset and AUC of 0.8830 and F1 score of 0.6836 on WCM dataset. Conclusion: We developed a deep learning framework using BERT models which provide an effective solution for prediction of MCI-to-AD progression using clinical note analysis.

Liang Yao, Jiazhen Peng, Chengsheng Mao et al.

Knowledge graphs play a vital role in numerous artificial intelligence tasks, yet they frequently face the issue of incompleteness. In this study, we explore utilizing Large Language Models (LLM) for knowledge graph completion. We consider triples in knowledge graphs as text sequences and introduce an innovative framework called Knowledge Graph LLM (KG-LLM) to model these triples. Our technique employs entity and relation descriptions of a triple as prompts and utilizes the response for predictions. Experiments on various benchmark knowledge graphs demonstrate that our method attains state-of-the-art performance in tasks such as triple classification and relation prediction. We also find that fine-tuning relatively smaller models (e.g., LLaMA-7B, ChatGLM-6B) outperforms recent ChatGPT and GPT-4.

Chengsheng Mao, Liang Yao, Yuan Luo

Acute kidney injury (AKI) is a common clinical syndrome characterized by a sudden episode of kidney failure or kidney damage within a few hours or a few days. Accurate early prediction of AKI for patients in ICU who are more likely than others to have AKI can enable timely interventions, and reduce the complications of AKI. Much of the clinical information relevant to AKI is captured in clinical notes that are largely unstructured text and requires advanced natural language processing (NLP) for useful information extraction. On the other hand, pre-trained contextual language models such as Bidirectional Encoder Representations from Transformers (BERT) have improved performances for many NLP tasks in general domain recently. However, few have explored BERT on disease-specific medical domain tasks such as AKI early prediction. In this paper, we try to apply BERT to specific diseases and present an AKI domain-specific pre-trained language model based on BERT (AKI-BERT) that could be used to mine the clinical notes for early prediction of AKI. AKI-BERT is a BERT model pre-trained on the clinical notes of patients having risks for AKI. Our experiments on Medical Information Mart for Intensive Care III (MIMIC-III) dataset demonstrate that AKI-BERT can yield performance improvements for early AKI prediction, thus expanding the utility of the BERT model from general clinical domain to disease-specific domain.

Yawei Li, Xin Wu, Ping Yang et al.

The recent development of imaging and sequencing technologies enables systematic advances in the clinical study of lung cancer. Meanwhile, the human mind is limited in effectively handling and fully utilizing the accumulation of such enormous amounts of data. Machine learning-based approaches play a critical role in integrating and analyzing these large and complex datasets, which have extensively characterized lung cancer through the use of different perspectives from these accrued data. In this article, we provide an overview of machine learning-based approaches that strengthen the varying aspects of lung cancer diagnosis and therapy, including early detection, auxiliary diagnosis, prognosis prediction and immunotherapy practice. Moreover, we highlight the challenges and opportunities for future applications of machine learning in lung cancer.

Yikuan Li, Hanyin Wang, Halid Yerebakan et al.

In this study, we investigated the ability of the large language model (LLM) to enhance healthcare data interoperability. We leveraged the LLM to convert clinical texts into their corresponding FHIR resources. Our experiments, conducted on 3,671 snippets of clinical text, demonstrated that the LLM not only streamlines the multi-step natural language processing and human calibration processes but also achieves an exceptional accuracy rate of over 90% in exact matches when compared to human annotations.

Ziqing Wang, Chengsheng Mao, Xiaole Wen et al.

Medical Multimodal Large Language Models (Med-MLLMs) have shown great promise in medical visual question answering (Med-VQA). However, when deployed in low-resource settings where abundant labeled data are unavailable, existing Med-MLLMs commonly fail due to their medical reasoning capability bottlenecks: (i) the intrinsic reasoning bottleneck that ignores the details from the medical image; (ii) the extrinsic reasoning bottleneck that fails to incorporate specialized medical knowledge. To address those limitations, we propose AMANDA, a training-free agentic framework that performs medical knowledge augmentation via LLM agents. Specifically, our intrinsic medical knowledge augmentation focuses on coarse-to-fine question decomposition for comprehensive diagnosis, while extrinsic medical knowledge augmentation grounds the reasoning process via biomedical knowledge graph retrieval. Extensive experiments across eight Med-VQA benchmarks demonstrate substantial improvements in both zero-shot and few-shot Med-VQA settings. The code is available at https://github.com/REAL-Lab-NU/AMANDA.

Chengsheng Mao, Liang Yao, Yuan Luo

Image representation is a fundamental task in computer vision. However, most of the existing approaches for image representation ignore the relations between images and consider each input image independently. Intuitively, relations between images can help to understand the images and maintain model consistency over related images, leading to better explainability. In this paper, we consider modeling the image-level relations to generate more informative image representations, and propose ImageGCN, an end-to-end graph convolutional network framework for inductive multi-relational image modeling. We apply ImageGCN to chest X-ray images where rich relational information is available for disease identification. Unlike previous image representation models, ImageGCN learns the representation of an image using both its original pixel features and its relationship with other images. Besides learning informative representations for images, ImageGCN can also be used for object detection in a weakly supervised manner. The experimental results on 3 open-source x-ray datasets, ChestX-ray14, CheXpert and MIMIC-CXR demonstrate that ImageGCN can outperform respective baselines in both disease identification and localization tasks and can achieve comparable and often better results than the state-of-the-art methods.

Yuan Luo, Song Han, Jingjing Liu

Machine learning is expected to enable the next Industrial Revolution. However, lacking standardized and automated assembly networks, ML faces significant challenges to meet ever-growing enterprise demands and empower broad industries. In the Perspective, we argue that ML needs to first complete its own Industrial Revolution, elaborate on how to best achieve its goals, and discuss new opportunities to enable rapid translation from ML's innovation frontier to mass production and utilization.

Xikun Jiang, He Lyu, Chenhao Ying et al.

With the increasingly widespread application of machine learning, how to strike a balance between protecting the privacy of data and algorithm parameters and ensuring the verifiability of machine learning has always been a challenge. This study explores the intersection of reinforcement learning and data privacy, specifically addressing the Multi-Armed Bandit (MAB) problem with the Upper Confidence Bound (UCB) algorithm. We introduce zkUCB, an innovative algorithm that employs the Zero-Knowledge Succinct Non-Interactive Argument of Knowledge (zk-SNARKs) to enhance UCB. zkUCB is carefully designed to safeguard the confidentiality of training data and algorithmic parameters, ensuring transparent UCB decision-making. Experiments highlight zkUCB's superior performance, attributing its enhanced reward to judicious quantization bit usage that reduces information entropy in the decision-making process. zkUCB's proof size and verification time scale linearly with the execution steps of zkUCB. This showcases zkUCB's adept balance between data security and operational efficiency. This approach contributes significantly to the ongoing discourse on reinforcing data privacy in complex decision-making processes, offering a promising solution for privacy-sensitive applications.

Chengsheng Mao, Yuan Luo

Graph neural network (GNN) is effective to model graphs for distributed representations of nodes and an entire graph. Recently, research on the expressive power of GNN attracted growing attention. A highly-expressive GNN has the ability to generate discriminative graph representations. However, in the end-to-end training process for a certain graph learning task, a highly-expressive GNN risks generating graph representations overfitting the training data for the target task, while losing information important for the model generalization. In this paper, we propose Distribution Preserving GNN (DP-GNN) - a GNN framework that can improve the generalizability of expressive GNN models by preserving several kinds of distribution information in graph representations and node representations. Besides the generalizability, by applying an expressive GNN backbone, DP-GNN can also have high expressive power. We evaluate the proposed DP-GNN framework on multiple benchmark datasets for graph classification tasks. The experimental results demonstrate that our model achieves state-of-the-art performances.

Alexander Cao, Diego Klabjan, Yuan Luo

Open-set recognition generalizes a classification task by classifying test samples as one of the known classes from training or "unknown." As novel cancer drug cocktails with improved treatment are continually discovered, predicting cancer treatments can naturally be formulated in terms of an open-set recognition problem. Drawbacks, due to modeling unknown samples during training, arise from straightforward implementations of prior work in healthcare open-set learning. Accordingly, we reframe the problem methodology and apply a recent existing Gaussian mixture variational autoencoder model, which achieves state-of-the-art results for image datasets, to breast cancer patient data. Not only do we obtain more accurate and robust classification results, with a 24.5% average F1 increase compared to a recent method, but we also reexamine open-set recognition in terms of deployability to a clinical setting.

Francisco Y Cai, Chengsheng Mao, Yuan Luo

Background: The increasing adoption of electronic health records (EHR) across the US has created troves of computable data, to which machine learning methods have been applied to extract useful insights. EHR data, represented as a three-dimensional analogue of a matrix (tensor), is decomposed into two-dimensional factors that can be interpreted as computational phenotypes. Methods: We apply constrained tensor factorization to derive computational phenotypes and predict mortality in cohorts of patients with breast, prostate, colorectal, or lung cancer in the Northwestern Medicine Enterprise Data Warehouse from 2000 to 2015. In our experiments, we examined using a supervised term in the factorization algorithm, filtering tensor co-occurrences by medical indication, and incorporating additional social determinants of health (SDOH) covariates in the factorization process. We evaluated the resulting computational phenotypes qualitatively and by assessing their ability to predict five-year mortality using the area under the curve (AUC) statistic. Results: Filtering by medical indication led to more concise and interpretable phenotypes. Mortality prediction performance (AUC) varied under the different experimental conditions and by cancer type (breast: 0.623 - 0.694, prostate: 0.603 - 0.750, colorectal: 0.523 - 0.641, and lung: 0.517 - 0.623). Generally, prediction performance improved with the use of a supervised term and the incorporation of SDOH covariates. Conclusion: Constrained tensor factorization, applied to sparse EHR data of patients with cancer, can discover computational phenotypes predictive of five-year mortality. The incorporation of SDOH variables into the factorization algorithm is an easy-to-implement and effective way to improve prediction performance.

Yu Deng, Jennifer A. Pacheco, Anh Chung et al.

Systemic lupus erythematosus (SLE) is a rare autoimmune disorder characterized by an unpredictable course of flares and remission with diverse manifestations. Lupus nephritis, one of the major disease manifestations of SLE for organ damage and mortality, is a key component of lupus classification criteria. Accurately identifying lupus nephritis in electronic health records (EHRs) would therefore benefit large cohort observational studies and clinical trials where characterization of the patient population is critical for recruitment, study design, and analysis. Lupus nephritis can be recognized through procedure codes and structured data, such as laboratory tests. However, other critical information documenting lupus nephritis, such as histologic reports from kidney biopsies and prior medical history narratives, require sophisticated text processing to mine information from pathology reports and clinical notes. In this study, we developed algorithms to identify lupus nephritis with and without natural language processing (NLP) using EHR data. We developed four algorithms: a rule-based algorithm using only structured data (baseline algorithm) and three algorithms using different NLP models. The three NLP models are based on regularized logistic regression and use different sets of features including positive mention of concept unique identifiers (CUIs), number of appearances of CUIs, and a mixture of three components respectively. The baseline algorithm and the best performed NLP algorithm were external validated on a dataset from Vanderbilt University Medical Center (VUMC). Our best performing NLP model incorporating features from both structured data, regular expression concepts, and mapped CUIs improved F measure in both the NMEDW (0.41 vs 0.79) and VUMC (0.62 vs 0.96) datasets compared to the baseline lupus nephritis algorithm.

Hanyin Wang, Yikuan Li, Andrew Naidech et al.

Background Sepsis is one of the most life-threatening circumstances for critically ill patients in the US, while a standardized criteria for sepsis identification is still under development. Disparities in social determinants of sepsis patients can interfere with the risk prediction performances using machine learning. Methods Disparities in social determinants, including race, gender, marital status, insurance types and languages, among patients identified by six available sepsis criteria were revealed by forest plots. Sixteen machine learning classifiers were trained to predict in-hospital mortality for sepsis patients. The performance of the trained model was tested on the entire randomly conducted test set and each sub-population built based on each of the following social determinants: race, gender, marital status, insurance type, and language. Results We analyzed a total of 11,791 critical care patients from the MIMIC-III database. Within the population identified by each sepsis identification method, significant differences were observed among sub-populations regarding race, marital status, insurance type, and language. On the 5,783 sepsis patients identified by the Sepsis-3 criteria statistically significant performance decreases for mortality prediction were observed when applying the trained machine learning model on Asian and Hispanic patients. With pairwise comparison, we detected performance discrepancies in mortality prediction between Asian and White patients, Asians and patients of other races, as well as English-speaking and Spanish-speaking patients. Conclusions Disparities in proportions of patients identified by various sepsis criteria were detected among the different social determinant groups. To achieve accurate diagnosis, a versatile diagnostic system for sepsis is needed to overcome the social determinant disparities of patients.

Yiming Li, Sanjiv J. Shah, Donna Arnett et al.

Hypertension is the leading global cause of cardiovascular disease and premature death. Distinct hypertension subtypes may vary in their prognoses and require different treatments. An individual's risk for hypertension is determined by genetic and environmental factors as well as their interactions. In this work, we studied 911 African Americans and 1,171 European Americans in the Hypertension Genetic Epidemiology Network (HyperGEN) cohort. We built hypertension subtype classification models using both environmental variables and sets of genetic features selected based on different criteria. The fitted prediction models provided insights into the genetic landscape of hypertension subtypes, which may aid personalized diagnosis and treatment of hypertension in the future.

Songzi Liu, Yuan Luo

Machine learning in medicine leverages the wealth of healthcare data to extract knowledge, facilitate clinical decision-making, and ultimately improve care delivery. However, ML models trained on datasets that lack demographic diversity could yield suboptimal performance when applied to the underrepresented populations (e.g. ethnic minorities, lower social-economic status), thus perpetuating health disparity. In this study, we evaluated four classifiers built to predict Hyperchloremia - a condition that often results from aggressive fluids administration in the ICU population - and compared their performance in racial, gender, and insurance subgroups. We observed that adding social determinants features in addition to the lab-based ones improved model performance on all patients. The subgroup testing yielded significantly different AUC scores in 40 out of the 44 model-subgroup, suggesting disparities when applying ML models to social determinants subgroups. We urge future researchers to design models that proactively adjust for potential biases and include subgroup reporting in their studies.

Jiyoung Shin, Yikuan Li, Yuan Luo

Sepsis is an important cause of mortality, especially in intensive care unit (ICU) patients. Developing novel methods to identify early mortality is critical for improving survival outcomes in sepsis patients. Using the MIMIC-III database, we integrated demographic data, physiological measurements and clinical notes. We built and applied several machine learning models to predict the risk of hospital mortality and 30-day mortality in sepsis patients. From the clinical notes, we generated clinically meaningful word representations and embeddings. Supervised learning classifiers and a deep learning architecture were used to construct prediction models. The configurations that utilized both structured and unstructured clinical features yielded competitive F-measure of 0.512. Our results showed that the approaches integrating both structured and unstructured clinical features can be effectively applied to assist clinicians in identifying the risk of mortality in sepsis patients upon admission to the ICU.

Yiqing Zhao, Yuan Luo

Delirium is a common acute onset brain dysfunction in the emergency setting and is associated with higher mortality. It is difficult to detect and monitor since its presentations and risk factors can be different depending on the underlying medical condition of patients. In our study, we aimed to identify subtypes within the delirium population and build subgroup-specific predictive models to detect delirium using Medical Information Mart for Intensive Care IV (MIMIC-IV) data. We showed that clusters exist within the delirium population. Differences in feature importance were also observed for subgroup-specific predictive models. Our work could recalibrate existing delirium prediction models for each delirium subgroup and improve the precision of delirium detection and monitoring for ICU or emergency department patients who had highly heterogeneous medical conditions.

Ye Xue, Diego Klabjan, Yuan Luo

Federated learning is a distributed machine learning paradigm where multiple data owners (clients) collaboratively train one machine learning model while keeping data on their own devices. The heterogeneity of client datasets is one of the most important challenges of federated learning algorithms. Studies have found performance reduction with standard federated algorithms, such as FedAvg, on non-IID data. Many existing works on handling non-IID data adopt the same aggregation framework as FedAvg and focus on improving model updates either on the server side or on clients. In this work, we tackle this challenge in a different view by introducing redistribution rounds that delay the aggregation. We perform experiments on multiple tasks and show that the proposed framework significantly improves the performance on non-IID data.

Adam Sandler, Diego Klabjan, Yuan Luo

We develop regularization methods to find flat minima while training deep neural networks. These minima generalize better than sharp minima, yielding models outperforming baselines on real-world test data (which may be distributed differently than the training data). Specifically, we propose a method of regularized optimization to reduce the spectral radius of the Hessian of the loss function. We also derive algorithms to efficiently optimize neural network models and prove that these algorithms almost surely converge. Furthermore, we demonstrate that our algorithm works effectively on applications in different domains, including healthcare. To show that our models generalize well, we introduced various methods for testing generalizability and found that our models outperform comparable baseline models on these tests.

Yuan Luo, Chengsheng Mao

Genetic pathways usually encode molecular mechanisms that can inform targeted interventions. It is often challenging for existing machine learning approaches to jointly model genetic pathways (higher-order features) and variants (atomic features), and present to clinicians interpretable models. In order to build more accurate and better interpretable machine learning models for genetic medicine, we introduce Pathway Augmented Nonnegative Tensor factorization for HighER-order feature learning (PANTHER). PANTHER selects informative genetic pathways that directly encode molecular mechanisms. We apply genetically motivated constrained tensor factorization to group pathways in a way that reflects molecular mechanism interactions. We then train a softmax classifier for disease types using the identified pathway groups. We evaluated PANTHER against multiple state-of-the-art constrained tensor/matrix factorization models, as well as group guided and Bayesian hierarchical models. PANTHER outperforms all state-of-the-art comparison models significantly (p<0.05). Our experiments on large scale Next Generation Sequencing (NGS) and whole-genome genotyping datasets also demonstrated wide applicability of PANTHER. We performed feature analysis in predicting disease types, which suggested insights and benefits of the identified pathway groups.

Chengsheng Mao, Liang Yao, Yuan Luo

Graph Neural Network (GNN) aggregates the neighborhood of each node into the node embedding and shows its powerful capability for graph representation learning. However, most existing GNN variants aggregate the neighborhood information in a fixed non-injective fashion, which may map different graphs or nodes to the same embedding, reducing the model expressiveness. We present a theoretical framework to design a continuous injective set function for neighborhood aggregation in GNN. Using the framework, we propose expressive GNN that aggregates the neighborhood of each node with a continuous injective set function, so that a GNN layer maps similar nodes with similar neighborhoods to similar embeddings, different nodes to different embeddings and the equivalent nodes or isomorphic graphs to the same embeddings. Moreover, the proposed expressive GNN can naturally learn expressive representations for graphs with continuous node attributes. We validate the proposed expressive GNN (ExpGNN) for graph classification on multiple benchmark datasets including simple graphs and attributed graphs. The experimental results demonstrate that our model achieves state-of-the-art performances on most of the benchmarks.

Yikuan Li, Hanyin Wang, Yuan Luo

Joint image-text embedding extracted from medical images and associated contextual reports is the bedrock for most biomedical vision-and-language (V+L) tasks, including medical visual question answering, clinical image-text retrieval, clinical report auto-generation. In this study, we adopt four pre-trained V+L models: LXMERT, VisualBERT, UNIER and PixelBERT to learn multimodal representation from MIMIC-CXR radiographs and associated reports. The extrinsic evaluation on OpenI dataset shows that in comparison to the pioneering CNN-RNN model, the joint embedding learned by pre-trained V+L models demonstrate performance improvement in the thoracic findings classification task. We conduct an ablation study to analyze the contribution of certain model components and validate the advantage of joint embedding over text-only embedding. We also visualize attention maps to illustrate the attention mechanism of V+L models.

Alexander Cao, Yuan Luo, Diego Klabjan

In inference, open-set classification is to either classify a sample into a known class from training or reject it as an unknown class. Existing deep open-set classifiers train explicit closed-set classifiers, in some cases disjointly utilizing reconstruction, which we find dilutes the latent representation's ability to distinguish unknown classes. In contrast, we train our model to cooperatively learn reconstruction and perform class-based clustering in the latent space. With this, our Gaussian mixture variational autoencoder (GMVAE) achieves more accurate and robust open-set classification results, with an average F1 improvement of 29.5%, through extensive experiments aided by analytical results.

Shaika Chowdhury, Chenwei Zhang, Philip S. Yu et al.

Distributed representations of medical concepts have been used to support downstream clinical tasks recently. Electronic Health Records (EHR) capture different aspects of patients' hospital encounters and serve as a rich source for augmenting clinical decision making by learning robust medical concept embeddings. However, the same medical concept can be recorded in different modalities (e.g., clinical notes, lab results)-with each capturing salient information unique to that modality-and a holistic representation calls for relevant feature ensemble from all information sources. We hypothesize that representations learned from heterogeneous data types would lead to performance enhancement on various clinical informatics and predictive modeling tasks. To this end, our proposed approach makes use of meta-embeddings, embeddings aggregated from learned embeddings. Firstly, modality-specific embeddings for each medical concept is learned with graph autoencoders. The ensemble of all the embeddings is then modeled as a meta-embedding learning problem to incorporate their correlating and complementary information through a joint reconstruction. Empirical results of our model on both quantitative and qualitative clinical evaluations have shown improvements over state-of-the-art embedding models, thus validating our hypothesis.

Adam Sandler, Diego Klabjan, Yuan Luo

We analyze large, multi-dimensional, sparse counting data sets, finding unsupervised groups to provide unique insights into genetic data. We create gene and biological pathway groups based on patients' variants to find common risk factors for four common types of cancer (breast, lung, prostate, and colorectal) and autism spectrum disorder. To accomplish this, we extend latent Dirichlet allocation to multiple dimensions and design distinct methods for hierarchical topic modeling. We find that our conditional hierarchical Bayesian Tucker decomposition models are more coherent than baseline models.

Shaika Chowdhury, Chenwei Zhang, Philip S. Yu et al.

Predicting patient mortality is an important and challenging problem in the healthcare domain, especially for intensive care unit (ICU) patients. Electronic health notes serve as a rich source for learning patient representations, that can facilitate effective risk assessment. However, a large portion of clinical notes are unstructured and also contain domain specific terminologies, from which we need to extract structured information. In this paper, we introduce an embedding framework to learn semantically-plausible distributed representations of clinical notes that exploits the semantic correspondence between the unstructured texts and their corresponding structured knowledge, known as semantic frame, in a hierarchical fashion. Our approach integrates text modeling and semantic correspondence learning into a single model that comprises 1) an unstructured embedding module that makes use of self-similarity matrix representations in order to inject structural regularities of different segments inherent in clinical texts to promote local coherence, 2) a structured embedding module to embed the semantic frames (e.g., UMLS semantic types) with deep ConvNet and 3) a hierarchical semantic correspondence module that embeds by enhancing the interactions between text-semantic frame embedding pairs at multiple levels (i.e., words, sentence, note). Evaluations on multiple embedding benchmarks on post discharge intensive care patient mortality prediction tasks demonstrate its effectiveness compared to approaches that do not exploit the semantic interactions between structured and unstructured information present in clinical notes.

Shaika Chowdhury, Chenwei Zhang, Philip S. Yu et al.

Distributed representations have been used to support downstream tasks in healthcare recently. Healthcare data (e.g., electronic health records) contain multiple modalities of data from heterogeneous sources that can provide complementary information, alongside an added dimension to learning personalized patient representations. To this end, in this paper we propose a novel unsupervised encoder-decoder model, namely Mixed Pooling Multi-View Attention Autoencoder (MPVAA), that generates patient representations encapsulating a holistic view of their medical profile. Specifically, by first learning personalized graph embeddings pertaining to each patient's heterogeneous healthcare data, it then integrates the non-linear relationships among them into a unified representation through multi-view attention mechanism. Additionally, a mixed pooling strategy is incorporated in the encoding step to learn diverse information specific to each data modality. Experiments conducted for multiple tasks demonstrate the effectiveness of the proposed model over the state-of-the-art representation learning methods in healthcare.

Liang Yao, Chengsheng Mao, Yuan Luo

Knowledge graphs are important resources for many artificial intelligence tasks but often suffer from incompleteness. In this work, we propose to use pre-trained language models for knowledge graph completion. We treat triples in knowledge graphs as textual sequences and propose a novel framework named Knowledge Graph Bidirectional Encoder Representations from Transformer (KG-BERT) to model these triples. Our method takes entity and relation descriptions of a triple as input and computes scoring function of the triple with the KG-BERT language model. Experimental results on multiple benchmark knowledge graphs show that our method can achieve state-of-the-art performance in triple classification, link prediction and relation prediction tasks.

Yisen Wang, Xingjun Ma, Zaiyi Chen et al.

Training accurate deep neural networks (DNNs) in the presence of noisy labels is an important and challenging task. Though a number of approaches have been proposed for learning with noisy labels, many open issues remain. In this paper, we show that DNN learning with Cross Entropy (CE) exhibits overfitting to noisy labels on some classes ("easy" classes), but more surprisingly, it also suffers from significant under learning on some other classes ("hard" classes). Intuitively, CE requires an extra term to facilitate learning of hard classes, and more importantly, this term should be noise tolerant, so as to avoid overfitting to noisy labels. Inspired by the symmetric KL-divergence, we propose the approach of \textbf{Symmetric cross entropy Learning} (SL), boosting CE symmetrically with a noise robust counterpart Reverse Cross Entropy (RCE). Our proposed SL approach simultaneously addresses both the under learning and overfitting problem of CE in the presence of noisy labels. We provide a theoretical analysis of SL and also empirically show, on a range of benchmark and real-world datasets, that SL outperforms state-of-the-art methods. We also show that SL can be easily incorporated into existing methods in order to further enhance their performance.

Ye Xue, Diego Klabjan, Yuan Luo

The problem of missing values in multivariable time series is a key challenge in many applications such as clinical data mining. Although many imputation methods show their effectiveness in many applications, few of them are designed to accommodate clinical multivariable time series. In this work, we propose a multiple imputation model that capture both cross-sectional information and temporal correlations. We integrate Gaussian processes with mixture models and introduce individualized mixing weights to handle the variance of predictive confidence of Gaussian process models. The proposed model is compared with several state-of-the-art imputation algorithms on both real-world and synthetic datasets. Experiments show that our best model can provide more accurate imputation than the benchmarks on all of our datasets.

Zhenxing Xu, Jingyuan Chou, Xi Sheryl Zhang et al.

Acute Kidney Injury (AKI) is a common clinical syndrome characterized by the rapid loss of kidney excretory function, which aggravates the clinical severity of other diseases in a large number of hospitalized patients. Accurate early prediction of AKI can enable in-time interventions and treatments. However, AKI is highly heterogeneous, thus identification of AKI sub-phenotypes can lead to an improved understanding of the disease pathophysiology and development of more targeted clinical interventions. This study used a memory network-based deep learning approach to discover AKI sub-phenotypes using structured and unstructured electronic health record (EHR) data of patients before AKI diagnosis. We leveraged a real world critical care EHR corpus including 37,486 ICU stays. Our approach identified three distinct sub-phenotypes: sub-phenotype I is with an average age of 63.03$ \pm 17.25 $ years, and is characterized by mild loss of kidney excretory function (Serum Creatinine (SCr) $1.55\pm 0.34$ mg/dL, estimated Glomerular Filtration Rate Test (eGFR) $107.65\pm 54.98$ mL/min/1.73$m^2$). These patients are more likely to develop stage I AKI. Sub-phenotype II is with average age 66.81$ \pm 10.43 $ years, and was characterized by severe loss of kidney excretory function (SCr $1.96\pm 0.49$ mg/dL, eGFR $82.19\pm 55.92$ mL/min/1.73$m^2$). These patients are more likely to develop stage III AKI. Sub-phenotype III is with average age 65.07$ \pm 11.32 $ years, and was characterized moderate loss of kidney excretory function and thus more likely to develop stage II AKI (SCr $1.69\pm 0.32$ mg/dL, eGFR $93.97\pm 56.53$ mL/min/1.73$m^2$). Both SCr and eGFR are significantly different across the three sub-phenotypes with statistical testing plus postdoc analysis, and the conclusion still holds after age adjustment.

Prakash Adekkanattu, Guoqian Jiang, Yuan Luo et al.

While natural language processing (NLP) of unstructured clinical narratives holds the potential for patient care and clinical research, portability of NLP approaches across multiple sites remains a major challenge. This study investigated the portability of an NLP system developed initially at the Department of Veterans Affairs (VA) to extract 27 key cardiac concepts from free-text or semi-structured echocardiograms from three academic medical centers: Weill Cornell Medicine, Mayo Clinic and Northwestern Medicine. While the NLP system showed high precision and recall measurements for four target concepts (aortic valve regurgitation, left atrium size at end systole, mitral valve regurgitation, tricuspid valve regurgitation) across all sites, we found moderate or poor results for the remaining concepts and the NLP system performance varied between individual sites.

Chengsheng Mao, Liang Yao, Yuan Luo

Laboratory testing and medication prescription are two of the most important routines in daily clinical practice. Developing an artificial intelligence system that can automatically make lab test imputations and medication recommendations can save costs on potentially redundant lab tests and inform physicians of a more effective prescription. We present an intelligent medical system (named MedGCN) that can automatically recommend the patients' medications based on their incomplete lab tests, and can even accurately estimate the lab values that have not been taken. In our system, we integrate the complex relations between multiple types of medical entities with their inherent features in a heterogeneous graph. Then we model the graph to learn a distributed representation for each entity in the graph based on graph convolutional networks (GCN). By the propagation of graph convolutional networks, the entity representations can incorporate multiple types of medical information that can benefit multiple medical tasks. Moreover, we introduce a cross regularization strategy to reduce overfitting for multi-task training by the interaction between the multiple tasks. In this study, we construct a graph to associate 4 types of medical entities, i.e., patients, encounters, lab tests, and medications, and applied a graph neural network to learn node embeddings for medication recommendation and lab test imputation. we validate our MedGCN model on two real-world datasets: NMEDW and MIMIC-III. The experimental results on both datasets demonstrate that our model can outperform the state-of-the-art in both tasks. We believe that our innovative system can provide a promising and reliable way to assist physicians to make medication prescriptions and to save costs on potentially redundant lab tests.

Yizhen Zhong, Luke Rasmussen, Yu Deng et al.

The automatic development of phenotype algorithms from Electronic Health Record data with machine learning (ML) techniques is of great interest given the current practice is very time-consuming and resource intensive. The extraction of design patterns from phenotype algorithms is essential to understand their rationale and standard, with great potential to automate the development process. In this pilot study, we perform network visualization on the design patterns and their associations with phenotypes and sites. We classify design patterns using the fragments from previously annotated phenotype algorithms as the ground truth. The classification performance is used as a proxy for coherence at the attribution level. The bag-of-words representation with knowledge-based features generated a good performance in the classification task (0.79 macro-f1 scores). Good classification accuracy with simple features demonstrated the attribution coherence and the feasibility of automatic identification of design patterns. Our results point to both the feasibility and challenges of automatic identification of phenotyping design patterns, which would power the automatic development of phenotype algorithms.

Yuan Luo, Peter Szolovits

This paper presents a Lisp architecture for a portable NLP system, termed LAPNLP, for processing clinical notes. LAPNLP integrates multiple standard, customized and in-house developed NLP tools. Our system facilitates portability across different institutions and data systems by incorporating an enriched Common Data Model (CDM) to standardize necessary data elements. It utilizes UMLS to perform domain adaptation when integrating generic domain NLP tools. It also features stand-off annotations that are specified by positional reference to the original document. We built an interval tree based search engine to efficiently query and retrieve the stand-off annotations by specifying positional requirements. We also developed a utility to convert an inline annotation format to stand-off annotations to enable the reuse of clinical text datasets with inline annotations. We experimented with our system on several NLP facilitated tasks including computational phenotyping for lymphoma patients and semantic relation extraction for clinical notes. These experiments showcased the broader applicability and utility of LAPNLP.

Yikuan Li, Liang Yao, Chengsheng Mao et al.

Acute kidney injury (AKI) in critically ill patients is associated with significant morbidity and mortality. Development of novel methods to identify patients with AKI earlier will allow for testing of novel strategies to prevent or reduce the complications of AKI. We developed data-driven prediction models to estimate the risk of new AKI onset. We generated models from clinical notes within the first 24 hours following intensive care unit (ICU) admission extracted from Medical Information Mart for Intensive Care III (MIMIC-III). From the clinical notes, we generated clinically meaningful word and concept representations and embeddings, respectively. Five supervised learning classifiers and knowledge-guided deep learning architecture were used to construct prediction models. The best configuration yielded a competitive AUC of 0.779. Our work suggests that natural language processing of clinical notes can be applied to assist clinicians in identifying the risk of incident AKI onset in critically ill patients upon admission to the ICU.

Guoqing Chao, Chengsheng Mao, Fei Wang et al.

ICU mortality risk prediction is a tough yet important task. On one hand, due to the complex temporal data collected, it is difficult to identify the effective features and interpret them easily; on the other hand, good prediction can help clinicians take timely actions to prevent the mortality. These correspond to the interpretability and accuracy problems. Most existing methods lack of the interpretability, but recently Subgraph Augmented Nonnegative Matrix Factorization (SANMF) has been successfully applied to time series data to provide a path to interpret the features well. Therefore, we adopted this approach as the backbone to analyze the patient data. One limitation of the raw SANMF method is its poor prediction ability due to its unsupervised nature. To deal with this problem, we proposed a supervised SANMF algorithm by integrating the logistic regression loss function into the NMF framework and solved it with an alternating optimization procedure. We used the simulation data to verify the effectiveness of this method, and then we applied it to ICU mortality risk prediction and demonstrated its superiority over other conventional supervised NMF methods.

Chengsheng Mao, Yiheng Pan, Zexian Zeng et al.

Thoracic diseases are very serious health problems that plague a large number of people. Chest X-ray is currently one of the most popular methods to diagnose thoracic diseases, playing an important role in the healthcare workflow. However, reading the chest X-ray images and giving an accurate diagnosis remain challenging tasks for expert radiologists. With the success of deep learning in computer vision, a growing number of deep neural network architectures were applied to chest X-ray image classification. However, most of the previous deep neural network classifiers were based on deterministic architectures which are usually very noise-sensitive and are likely to aggravate the overfitting issue. In this paper, to make a deep architecture more robust to noise and to reduce overfitting, we propose using deep generative classifiers to automatically diagnose thorax diseases from the chest X-ray images. Unlike the traditional deterministic classifier, a deep generative classifier has a distribution middle layer in the deep neural network. A sampling layer then draws a random sample from the distribution layer and input it to the following layer for classification. The classifier is generative because the class label is generated from samples of a related distribution. Through training the model with a certain amount of randomness, the deep generative classifiers are expected to be robust to noise and can reduce overfitting and then achieve good performances. We implemented our deep generative classifiers based on a number of well-known deterministic neural network architectures, and tested our models on the chest X-ray14 dataset. The results demonstrated the superiority of deep generative classifiers compared with the corresponding deep deterministic classifiers.

Chengsheng Mao, Liang Yao, Yuan Luo

In open set learning, a model must be able to generalize to novel classes when it encounters a sample that does not belong to any of the classes it has seen before. Open set learning poses a realistic learning scenario that is receiving growing attention. Existing studies on open set learning mainly focused on detecting novel classes, but few studies tried to model them for differentiating novel classes. In this paper, we recognize that novel classes should be different from each other, and propose distribution networks for open set learning that can model different novel classes based on probability distributions. We hypothesize that, through a certain mapping, samples from different classes with the same classification criterion should follow different probability distributions from the same distribution family. A deep neural network is learned to map the samples in the original feature space to a latent space where the distributions of known classes can be jointly learned with the network. We additionally propose a distribution parameter transfer and updating strategy for novel class modeling when a novel class is detected in the latent space. By novel class modeling, the detected novel classes can serve as known classes to the subsequent classification. Our experimental results on image datasets MNIST and CIFAR10 show that the distribution networks can detect novel classes accurately, and model them well for the subsequent classification tasks.

Liang Yao, Chengsheng Mao, Yuan Luo

Text classification is an important and classical problem in natural language processing. There have been a number of studies that applied convolutional neural networks (convolution on regular grid, e.g., sequence) to classification. However, only a limited number of studies have explored the more flexible graph convolutional neural networks (convolution on non-grid, e.g., arbitrary graph) for the task. In this work, we propose to use graph convolutional networks for text classification. We build a single text graph for a corpus based on word co-occurrence and document word relations, then learn a Text Graph Convolutional Network (Text GCN) for the corpus. Our Text GCN is initialized with one-hot representation for word and document, it then jointly learns the embeddings for both words and documents, as supervised by the known class labels for documents. Our experimental results on multiple benchmark datasets demonstrate that a vanilla Text GCN without any external word embeddings or knowledge outperforms state-of-the-art methods for text classification. On the other hand, Text GCN also learns predictive word and document embeddings. In addition, experimental results show that the improvement of Text GCN over state-of-the-art comparison methods become more prominent as we lower the percentage of training data, suggesting the robustness of Text GCN to less training data in text classification.

Liang Yao, Chengsheng Mao, Yuan Luo

Clinical text classification is an important problem in medical natural language processing. Existing studies have conventionally focused on rules or knowledge sources-based feature engineering, but only a few have exploited effective feature learning capability of deep learning methods. In this study, we propose a novel approach which combines rule-based features and knowledge-guided deep learning techniques for effective disease classification. Critical Steps of our method include identifying trigger phrases, predicting classes with very few examples using trigger phrases and training a convolutional neural network with word embeddings and Unified Medical Language System (UMLS) entity embeddings. We evaluated our method on the 2008 Integrating Informatics with Biology and the Bedside (i2b2) obesity challenge. The results show that our method outperforms the state of the art methods.

Himanshu Sharma, Chengsheng Mao, Yizhen Zhang et al.

This paper presents a portable phenotyping system that is capable of integrating both rule-based and statistical machine learning based approaches. Our system utilizes UMLS to extract clinically relevant features from the unstructured text and then facilitates portability across different institutions and data systems by incorporating OHDSI's OMOP Common Data Model (CDM) to standardize necessary data elements. Our system can also store the key components of rule-based systems (e.g., regular expression matches) in the format of OMOP CDM, thus enabling the reuse, adaptation and extension of many existing rule-based clinical NLP systems. We experimented with our system on the corpus from i2b2's Obesity Challenge as a pilot study. Our system facilitates portable phenotyping of obesity and its 15 comorbidities based on the unstructured patient discharge summaries, while achieving a performance that often ranked among the top 10 of the challenge participants. This standardization enables a consistent application of numerous rule-based and machine learning based classification techniques downstream.

Zexian Zeng, Yu Deng, Xiaoyu Li et al.

This article reviews recent advances in applying natural language processing (NLP) to Electronic Health Records (EHRs) for computational phenotyping. NLP-based computational phenotyping has numerous applications including diagnosis categorization, novel phenotype discovery, clinical trial screening, pharmacogenomics, drug-drug interaction (DDI) and adverse drug event (ADE) detection, as well as genome-wide and phenome-wide association studies. Significant progress has been made in algorithm development and resource construction for computational phenotyping. Among the surveyed methods, well-designed keyword search and rule-based systems often achieve good performance. However, the construction of keyword and rule lists requires significant manual effort, which is difficult to scale. Supervised machine learning models have been favored because they are capable of acquiring both classification patterns and structures from data. Recently, deep learning and unsupervised learning have received growing attention, with the former favored for its performance and the latter for its ability to find novel phenotypes. Integrating heterogeneous data sources have become increasingly important and have shown promise in improving model performance. Often better performance is achieved by combining multiple modalities of information. Despite these many advances, challenges and opportunities remain for NLP-based computational phenotyping, including better model interpretability and generalizability, and proper characterization of feature relations in clinical narratives

Zexian Zeng, Ankita Roy, Xiaoyu Li et al.

Accurately identifying distant recurrences in breast cancer from the Electronic Health Records (EHR) is important for both clinical care and secondary analysis. Although multiple applications have been developed for computational phenotyping in breast cancer, distant recurrence identification still relies heavily on manual chart review. In this study, we aim to develop a model that identifies distant recurrences in breast cancer using clinical narratives and structured data from EHR. We apply MetaMap to extract features from clinical narratives and also retrieve structured clinical data from EHR. Using these features, we train a support vector machine model to identify distant recurrences in breast cancer patients. We train the model using 1,396 double-annotated subjects and validate the model using 599 double-annotated subjects. In addition, we validate the model on a set of 4,904 single-annotated subjects as a generalization test. We obtained a high area under curve (AUC) score of 0.92 (SD=0.01) in the cross-validation using the training dataset, then obtained AUC scores of 0.95 and 0.93 in the held-out test and generalization test using 599 and 4,904 samples respectively. Our model can accurately and efficiently identify distant recurrences in breast cancer by combining features extracted from unstructured clinical narratives and structured clinical data.

Chengsheng Mao, Yuan Zhao, Mengxin Sun et al.

Privacy is a major concern in sharing human subject data to researchers for secondary analyses. A simple binary consent (opt-in or not) may significantly reduce the amount of sharable data, since many patients might only be concerned about a few sensitive medical conditions rather than the entire medical records. We propose event-level privacy protection, and develop a feature ablation method to protect event-level privacy in electronic medical records. Using a list of 13 sensitive diagnoses, we evaluate the feasibility and the efficacy of the proposed method. As feature ablation progresses, the identifiability of a sensitive medical condition decreases with varying speeds on different diseases. We find that these sensitive diagnoses can be divided into 3 categories: (1) 5 diseases have fast declining identifiability (AUC below 0.6 with less than 400 features excluded); (2) 7 diseases with progressively declining identifiability (AUC below 0.7 with between 200 and 700 features excluded); and (3) 1 disease with slowly declining identifiability (AUC above 0.7 with 1000 features excluded). The fact that the majority (12 out of 13) of the sensitive diseases fall into the first two categories suggests the potential of the proposed feature ablation method as a solution for event-level record privacy protection.

Xi Sheryl Zhang, Lifang He, Kun Chen et al.

Parkinson's Disease (PD) is one of the most prevalent neurodegenerative diseases that affects tens of millions of Americans. PD is highly progressive and heterogeneous. Quite a few studies have been conducted in recent years on predictive or disease progression modeling of PD using clinical and biomarkers data. Neuroimaging, as another important information source for neurodegenerative disease, has also arisen considerable interests from the PD community. In this paper, we propose a deep learning method based on Graph Convolutional Networks (GCN) for fusing multiple modalities of brain images in relationship prediction which is useful for distinguishing PD cases from controls. On Parkinson's Progression Markers Initiative (PPMI) cohort, our approach achieved $0.9537\pm 0.0587$ AUC, compared with $0.6443\pm 0.0223$ AUC achieved by traditional approaches such as PCA.