MetaPred: Meta-Learning for Clinical Risk Prediction with Limited Patient Electronic Health Records

Deployment and validation of the CLL treatment infection model adjoined to an EHR system

Research algorithms are seldom externally validated or integrated into clinical practice, leaving unknown challenges in deployment. In such efforts, one needs to address challenges related to data harmonization, the performance of an algorithm in unforeseen missingness, automation and monitoring of predictions, and legal frameworks. We here describe the deployment of a high-dimensional data-driven decision support model into an EHR and derive practical guidelines informed by this deployment that includes the necessary processes, stakeholders and design requirements for a successful deployment. For this, we describe our deployment of the chronic lymphocytic leukemia (CLL) treatment infection model (CLL-TIM) as a stand-alone platform adjoined to an EPIC-based Danish Electronic Health Record (EHR), with the presentation of personalized predictions in a clinical context. CLL-TIM is an 84-variable data-driven prognostic model utilizing 7-year medical patient records and predicts the 2-year risk composite outcome of infection and/or treatment post-CLL diagnosis. As an independent validation cohort for this deployment, we used a retrospective population-based cohort of patients diagnosed with CLL from 2018 onwards (n = 1480). Unexpectedly high levels of missingness for key CLL-TIM variables were exhibited upon deployment. High dimensionality, with the handling of missingness, and predictive confidence were critical design elements that enabled trustworthy predictions and thus serves as a priority for prognostic models seeking deployment in new EHRs. Our setup for deployment, including automation and monitoring into EHR that meets Medical Device Regulations, may be used as step-by-step guidelines for others aiming at designing and deploying research algorithms into clinical practice.

Meta Transfer of Self-Supervised Knowledge: Foundation Model in Action for Post-Traumatic Epilepsy Prediction

Despite the impressive advancements achieved using deep-learning for functional brain activity analysis, the heterogeneity of functional patterns and scarcity of imaging data still pose challenges in tasks such as prediction of future onset of Post-Traumatic Epilepsy (PTE) from data acquired shortly after traumatic brain injury (TBI). Foundation models pre-trained on separate large-scale datasets can improve the performance from scarce and heterogeneous datasets. For functional Magnetic Resonance Imaging (fMRI), while data may be abundantly available from healthy controls, clinical data is often scarce, limiting the ability of foundation models to identify clinically-relevant features. We overcome this limitation by introducing a novel training strategy for our foundation model by integrating meta-learning with self-supervised learning to improve the generalization from normal to clinical features. In this way we enable generalization to other downstream clinical tasks, in our case prediction of PTE. To achieve this, we perform self-supervised training on the control dataset to focus on inherent features that are not limited to a particular supervised task while applying meta-learning, which strongly improves the model's generalizability using bi-level optimization. Through experiments on neurological disorder classification tasks, we demonstrate that the proposed strategy significantly improves task performance on small-scale clinical datasets. To explore the generalizability of the foundation model in downstream applications, we then apply the model to an unseen TBI dataset for prediction of PTE using zero-shot learning. Results further demonstrated the enhanced generalizability of our foundation model.

MetaGeo: A General Framework for Social User Geolocation Identification With Few-Shot Learning

Identifying the geolocation of social media users is an important problem in a wide range of applications, spanning from disease outbreaks, emergency detection, local event recommendation, to fake news localization, online marketing planning, and even crime control and prevention. Researchers have attempted to propose various models by combining different sources of information, including text, social relation, and contextual data, which indeed has achieved promising results. However, existing approaches still suffer from certain constraints, such as: 1) a very few samples are available and 2) prediction models are not easy to be generalized for users from new regions-which are challenges that motivate our study. In this article, we propose a general framework for identifying user geolocation-MetaGeo, which is a meta-learning-based approach, learning the prior distribution of the geolocation task in order to quickly adapt the prediction toward users from new locations. Different from typical meta-learning settings that only learn a new concept from few-shot samples, MetaGeo improves the geolocation prediction with conventional settings by ensembling numerous mini-tasks. In addition, MetaGeo incorporates probabilistic inference to alleviate two issues inherent in training with few samples: location uncertainty and task ambiguity. To demonstrate the effectiveness of MetaGeo, we conduct extensive experimental evaluations on three real-world datasets and compare the performance with several state-of-the-art benchmark models. The results demonstrate the superiority of MetaGeo in both the settings where the predicted locations/regions are known or have not been seen during training.

A Few-shot learning approach for Monkeypox recognition from a cross-domain perspective

Monkeypox is a zoonotic infectious skin disease initially endemic in Africa only. However, some countries are now beginning to report cases of apparent community transmission. In Computer Aided Diagnosis, deep learning has gained substantial improvement over traditional methods. Commonly, training a supervised deep model requires a large number of labeled samples. However, the collection and annotation of new disease images such as human monkeypox are time-consuming and expensive. Thus, we introduce a few-shot learning based approach for the recognition of human monkeypox in images. It requires merely a small number of training samples. In particular, it is a novel framework built with a normal backbone and auxiliary backbones. They are co-trained with Self-supervised Learning and Cross-domain Adaption techniques. The self-supervision penalty is used to help the auxiliary backbones effectively learn priors from source domain. The combined features across different domains are unified through a power transform layer. Extensive experiments are conducted on a task of recognizing chickenpox, measles, and human monkeypox diseases in a three-way few-shot manner. The results demonstrate that our method outperforms mainstream few-shot learning algorithms such as meta-learning based and fine-tuning based methods.

Deep learning prediction models based on EHR trajectories: A systematic review

BACKGROUND

Electronic health records (EHRs) are generated at an ever-increasing rate. EHR trajectories, the temporal aspect of health records, facilitate predicting patients' future health-related risks. It enables healthcare systems to increase the quality of care through early identification and primary prevention. Deep learning techniques have shown great capacity for analyzing complex data and have been successful for prediction tasks using complex EHR trajectories. This systematic review aims to analyze recent studies to identify challenges, knowledge gaps, and ongoing research directions.

METHODS

For this systematic review, we searched Scopus, PubMed, IEEE Xplore, and ACM databases from Jan 2016 to April 2022 using search terms centered around EHR, deep learning, and trajectories. Then the selected papers were analyzed according to publication characteristics, objectives, and their solutions regarding existing challenges, such as the model's capacity to deal with intricate data dependencies, data insufficiency, and explainability.

RESULTS

After removing duplicates and out-of-scope papers, 63 papers were selected, which showed rapid growth in the number of research in recent years. Predicting all diseases in the next visit and the onset of cardiovascular diseases were the most common targets. Different contextual and non-contextual representation learning methods are employed to retrieve important information from the sequence of EHR trajectories. Recurrent neural networks and the time-aware attention mechanism for modeling long-term dependencies, self-attentions, convolutional neural networks, graphs for representing inner visit relations, and attention scores for explainability were frequently used among the reviewed publications.

CONCLUSIONS

This systematic review demonstrated how recent breakthroughs in deep learning methods have facilitated the modeling of EHR trajectories. Research on improving the ability of graph neural networks, attention mechanisms, and cross-modal learning to analyze intricate dependencies among EHRs has shown good progress. There is a need to increase the number of publicly available EHR trajectory datasets to allow for easier comparison among different models. Also, very few developed models can handle all aspects of EHR trajectory data.

Clinical Risk Prediction Models with Meta-Learning Prototypes of Patient Heterogeneity

Hospitalized patients sometimes have complex health conditions, such as multiple diseases, underlying diseases, and complications. The heterogeneous patient conditions may have various representations. A generalized model ignores the differences among heterogeneous patients, and personalized models, even with transfer learning, are still limited to the small amount of training data and the repeated training process. Meta-learning provides a solution for training similar patients based on few-shot learning; however, cannot address common cross-domain patients. Inspired by prototypical networks [1], we proposed a meta-prototype for Electronic Health Records (EHR), a meta-learning-based model with flexible prototypes representing the heterogeneity in patients. We apply this technique to cardiovascular diseases in MIMIC-III and compare it against a set of benchmark models, and demonstrate its ability to address heterogeneous patient health conditions and improve the model performances from 1.2% to 11.9% on different metrics and prediction tasks.Clinical relevance- Developing an adaptive EHR risk prediction model for outcomes-driven phenotyping of heterogeneous patient health conditions.

Predicting Progression from Normal to MCI and from MCI to AD Using Clinical Variables in the National Alzheimer's Coordinating Center Uniform Data Set Version 3: Application of Machine Learning Models and a Probability Calculator

Clinical trials are increasingly focused on pre-manifest and early Alzheimer's disease (AD). Accurately predicting clinical progressions from normal to MCI or from MCI to dementia/AD versus non-progression is challenging. Accurate identification of symptomatic progressors is important to avoid unnecessary treatment and improve trial efficiency. Due to large inter-individual variability, biomarker positivity and comorbidity information are often insufficient to identify those destined to have symptomatic progressions. Using only clinical variables, we aimed to predict clinical progressions, estimating probabilities of progressions with a small set of variables selected by machine learning approaches. This work updates our previous work that was applied to the National Alzheimer's Coordinating Center (NACC) Uniform Data Set Version 2 (V2), by using the most recent version (V3) with additional analyses. We generated a user-friendly conversion probability calculator which can be used for effectively pre-screening trial participants.

MR-KPA: medication recommendation by combining knowledge-enhanced pre-training with a deep adversarial network

BACKGROUND

Medication recommendation based on electronic medical record (EMR) is a research hot spot in smart healthcare. For developing computational medication recommendation methods based on EMR, an important challenge is the lack of a large number of longitudinal EMR data with time correlation. Faced with this challenge, this paper proposes a new EMR-based medication recommendation model called MR-KPA, which combines knowledge-enhanced pre-training with the deep adversarial network to improve medication recommendation from both feature representation and the fine-tuning process. Firstly, a knowledge-enhanced pre-training visit model is proposed to realize domain knowledge-based external feature fusion and pre-training-based internal feature mining for improving the feature representation. Secondly, a medication recommendation model based on the deep adversarial network is developed to optimize the fine-tuning process of pre-training visit model and alleviate over-fitting of model caused by the task gap between pre-training and recommendation.

RESULT

The experimental results on EMRs from medical and health institutions in Hainan Province, China show that the proposed MR-KPA model can effectively improve the accuracy of medication recommendation on small-scale longitudinal EMR data compared with existing representative methods.

CONCLUSION

The advantages of the proposed MR-KPA are mainly attributed to knowledge enhancement based on ontology embedding, the pre-training visit model and adversarial training. Each of these three optimizations is very effective for improving the capability of medication recommendation on small-scale longitudinal EMR data, and the pre-training visit model has the most significant improvement effect. These three optimizations are also complementary, and their integration makes the proposed MR-KPA model achieve the best recommendation effect.

Uncovering interpretable potential confounders in electronic medical records

Randomized clinical trials (RCT) are the gold standard for informing treatment decisions. Observational studies are often plagued by selection bias, and expert-selected covariates may insufficiently adjust for confounding. We explore how unstructured clinical text can be used to reduce selection bias and improve medical practice. We develop a framework based on natural language processing to uncover interpretable potential confounders from text. We validate our method by comparing the estimated hazard ratio (HR) with and without the confounders against established RCTs. We apply our method to four cohorts built from localized prostate and lung cancer datasets from the Stanford Cancer Institute and show that our method shifts the HR estimate towards the RCT results. The uncovered terms can also be interpreted by oncologists for clinical insights. We present this proof-of-concept study to enable more credible causal inference using observational data, uncover meaningful insights from clinical text, and inform high-stakes medical decisions.

Toward attention-based learning to predict the risk of brain degeneration with multimodal medical data

Introduction

Brain degeneration is commonly caused by some chronic diseases, such as Alzheimer's disease (AD) and diabetes mellitus (DM). The risk prediction of brain degeneration aims to forecast the situation of disease progression of patients in the near future based on their historical health records. It is beneficial for patients to make an accurate clinical diagnosis and early prevention of disease. Current risk predictions of brain degeneration mainly rely on single-modality medical data, such as Electronic Health Records (EHR) or magnetic resonance imaging (MRI). However, only leveraging EHR or MRI data for the pertinent and accurate prediction is insufficient because of single-modality information (e.g., pixel or volume information of image data or clinical context information of non-image data).

Methods

Several deep learning-based methods have used multimodal data to predict the risks of specified diseases. However, most of them simply integrate different modalities in an early, intermediate, or late fusion structure and do not care about the intra-modal and intermodal dependencies. A lack of these dependencies would lead to sub-optimal prediction performance. Thus, we propose an encoder-decoder framework for better risk prediction of brain degeneration by using MRI and EHR. An encoder module is one of the key components and mainly focuses on feature extraction of input data. Specifically, we introduce an encoder module, which integrates intra-modal and inter-modal dependencies with the spatial-temporal attention and cross-attention mechanism. The corresponding decoder module is another key component and mainly parses the features from the encoder. In the decoder module, a disease-oriented module is used to extract the most relevant disease representation features. We take advantage of a multi-head attention module followed by a fully connected layer to produce the predicted results.

Results

As different types of AD and DM influence the nature and severity of brain degeneration, we evaluate the proposed method for three-class prediction of AD and three-class prediction of DM. Our results show that the proposed method with integrated MRI and EHR data achieves an accuracy of 0.859 and 0.899 for the risk prediction of AD and DM, respectively.

Discussion

The prediction performance is significantly better than the benchmarks, including MRI-only, EHR-only, and state-of-the-art multimodal fusion methods.

Natural language processing in clinical neuroscience and psychiatry: A review

Natural language processing (NLP) is rapidly becoming an important topic in the medical community. The ability to automatically analyze any type of medical document could be the key factor to fully exploit the data it contains. Cutting-edge artificial intelligence (AI) architectures, particularly machine learning and deep learning, have begun to be applied to this topic and have yielded promising results. We conducted a literature search for 1,024 papers that used NLP technology in neuroscience and psychiatry from 2010 to early 2022. After a selection process, 115 papers were evaluated. Each publication was classified into one of three categories: information extraction, classification, and data inference. Automated understanding of clinical reports in electronic health records has the potential to improve healthcare delivery. Overall, the performance of NLP applications is high, with an average F1-score and AUC above 85%. We also derived a composite measure in the form of Z-scores to better compare the performance of NLP models and their different classes as a whole. No statistical differences were found in the unbiased comparison. Strong asymmetry between English and non-English models, difficulty in obtaining high-quality annotated data, and train biases causing low generalizability are the main limitations. This review suggests that NLP could be an effective tool to help clinicians gain insights from medical reports, clinical research forms, and more, making NLP an effective tool to improve the quality of healthcare services.

Deep learning for temporal data representation in electronic health records: A systematic review of challenges and methodologies

OBJECTIVE

Temporal electronic health records (EHRs) contain a wealth of information for secondary uses, such as clinical events prediction and chronic disease management. However, challenges exist for temporal data representation. We therefore sought to identify these challenges and evaluate novel methodologies for addressing them through a systematic examination of deep learning solutions.

METHODS

We searched five databases (PubMed, Embase, the Institute of Electrical and Electronics Engineers [IEEE] Xplore Digital Library, the Association for Computing Machinery [ACM] Digital Library, and Web of Science) complemented with hand-searching in several prestigious computer science conference proceedings. We sought articles that reported deep learning methodologies on temporal data representation in structured EHR data from January 1, 2010, to August 30, 2020. We summarized and analyzed the selected articles from three perspectives: nature of time series, methodology, and model implementation.

RESULTS

We included 98 articles related to temporal data representation using deep learning. Four major challenges were identified, including data irregularity, heterogeneity, sparsity, and model opacity. We then studied how deep learning techniques were applied to address these challenges. Finally, we discuss some open challenges arising from deep learning.

CONCLUSION

Temporal EHR data present several major challenges for clinical prediction modeling and data utilization. To some extent, current deep learning solutions can address these challenges. Future studies may consider designing comprehensive and integrated solutions. Moreover, researchers should incorporate clinical domain knowledge into study designs and enhance model interpretability to facilitate clinical implementation.

Time-Aware Multi-Type Data Fusion Representation Learning Framework for Risk Prediction of Cardiovascular Diseases

Predicting the future risk of cardiovascular diseases from the historical Electronic Health Records (EHRs) is a significant research task in personalized healthcare fields. In recent years, many deep neural network-based methods have emerged, which model patient disease progression by capturing the temporal patterns in sequential visit data. However, existing methods usually cannot effectively integrate the features of heterogeneous clinical data, and do not fully consider the impact of patients age and irregular time interval between consecutive medical records on the patients disease development. To address these challenges, we propose a Time-Aware Multi-type Data fUsion Representation learning framework (TAMDUR) for CVDs risk prediction. In this framework, we design a time-aware decay function, which is based on the patients age and the elapsed time between visits, to model the disease progression pattern. A parallel combination of Bi LSTM and CNN is constructed to respectively learn the temporal and non-temporal features from various types of clinical data. Finally, a multi-type data fusion representation layer based on self-attention is utilized to integrate various features and their correlations to obtain the final patient representation. We evaluate our model on a real medical dataset, and the experimental results demonstrate that TAMDUR outperforms the state-of-the-art approaches.

Stroke Risk Prediction with Hybrid Deep Transfer Learning Framework

Stroke has become a leading cause of death and long-term disability in the world, and there is no effective treatment.Deep learning-based approaches have the potential to outperform existing stroke risk prediction models, they rely on large well-labeled data. Due to the strict privacy protection policy in health-care systems, stroke data is usually distributed among different hospitals in small pieces. In addition, the positive and negative instances of such data are extremely imbalanced. Transfer learning solves small data issue by exploiting the knowledge of a correlated domain, especially when multiple source are available.In this work, we propose a novel Hybrid Deep Transfer Learning-based Stroke Risk Prediction (HDTL-SRP) scheme to exploit the knowledge structure from multiple correlated sources (i.e.,external stroke data, chronic diseases data, such as hypertension and diabetes). The proposed framework has been extensively tested in synthetic and real-world scenarios, and it outperforms the state-of-the-art stroke risk prediction models. It also shows the potential of real-world deployment among multiple hospitals aided with 5G/B5G infrastructures.

Disease Prediction via Graph Neural Networks

With the increasingly available electronic medical records (EMRs), disease prediction has recently gained immense research attention, where an accurate classifier needs to be trained to map the input prediction signals (e.g., symptoms, patient demographics, etc.) to the estimated diseases for each patient. However, existing machine learning-based solutions heavily rely on abundant manually labeled EMR training data to ensure satisfactory prediction results, impeding their performance in the existence of rare diseases that are subject to severe data scarcity. For each rare disease, the limited EMR data can hardly offer sufficient information for a model to correctly distinguish its identity from other diseases with similar clinical symptoms. Furthermore, most existing disease prediction approaches are based on the sequential EMRs collected for every patient and are unable to handle new patients without historical EMRs, reducing their real-life practicality. In this paper, we introduce an innovative model based on Graph Neural Networks (GNNs) for disease prediction, which utilizes external knowledge bases to augment the insufficient EMR data, and learns highly representative node embeddings for patients, diseases and symptoms from the medical concept graph and patient record graph respectively constructed from the medical knowledge base and EMRs. By aggregating information from directly connected neighbor nodes, the proposed neural graph encoder can effectively generate embeddings that capture knowledge from both data sources, and is able to inductively infer the embeddings for a new patient based on the symptoms reported in her/his EMRs to allow for accurate prediction on both general diseases and rare diseases. Extensive experiments on a real-world EMR dataset have demonstrated the state-of-the-art performance of our proposed model.

A causal framework to determine the effectiveness of dynamic quarantine policy to mitigate COVID-19

Since the start of the pandemic caused by the novel coronavirus, COVID-19, more than 106 million people have been infected and global deaths have surpassed 2.4 million. In Chile, the government restricted the activities and movement of people, organizations, and companies, under the concept of dynamic quarantine across municipalities for a predefined period of time. Chile is an interesting context to study because reports to have a higher quantity of infections per million people as well as a higher number of polymerize chain reaction (PCR) tests per million people. The higher testing rate means that Chile has good measurement of the contagious compared to other countries. Further, the heterogeneity of the social, economic, and demographic variables collected of each Chilean municipality provides a robust set of control data to better explain the contagious rate for each city. In this paper, we propose a framework to determine the effectiveness of the dynamic quarantine policy by analyzing different causal models (meta-learners and causal forest) including a time series pattern related to effective reproductive number. Additionally, we test the ability of the proposed framework to understand and explain the spread over benchmark traditional models and to interpret the Shapley Additive Explanations (SHAP) plots. The conclusions derived from the proposed framework provide important scientific information for government policymakers in disease control strategies, not only to analyze COVID-19 but to have a better model to determine social interventions for future outbreaks.

Deep representation learning of patient data from Electronic Health Records (EHR): A systematic review

OBJECTIVES

Patient representation learning refers to learning a dense mathematical representation of a patient that encodes meaningful information from Electronic Health Records (EHRs). This is generally performed using advanced deep learning methods. This study presents a systematic review of this field and provides both qualitative and quantitative analyses from a methodological perspective.

METHODS

We identified studies developing patient representations from EHRs with deep learning methods from MEDLINE, EMBASE, Scopus, the Association for Computing Machinery (ACM) Digital Library, and the Institute of Electrical and Electronics Engineers (IEEE) Xplore Digital Library. After screening 363 articles, 49 papers were included for a comprehensive data collection.

RESULTS

Publications developing patient representations almost doubled each year from 2015 until 2019. We noticed a typical workflow starting with feeding raw data, applying deep learning models, and ending with clinical outcome predictions as evaluations of the learned representations. Specifically, learning representations from structured EHR data was dominant (37 out of 49 studies). Recurrent Neural Networks were widely applied as the deep learning architecture (Long short-term memory: 13 studies, Gated recurrent unit: 11 studies). Learning was mainly performed in a supervised manner (30 studies) optimized with cross-entropy loss. Disease prediction was the most common application and evaluation (31 studies). Benchmark datasets were mostly unavailable (28 studies) due to privacy concerns of EHR data, and code availability was assured in 20 studies.

DISCUSSION & CONCLUSION

The existing predictive models mainly focus on the prediction of single diseases, rather than considering the complex mechanisms of patients from a holistic review. We show the importance and feasibility of learning comprehensive representations of patient EHR data through a systematic review. Advances in patient representation learning techniques will be essential for powering patient-level EHR analyses. Future work will still be devoted to leveraging the richness and potential of available EHR data. Reproducibility and transparency of reported results will hopefully improve. Knowledge distillation and advanced learning techniques will be exploited to assist the capability of learning patient representation further.

Multi-task Learning via Adaptation to Similar Tasks for Mortality Prediction of Diverse Rare Diseases

The mortality prediction of diverse rare diseases using electronic health record (EHR) data is a crucial task for intelligent healthcare. However, data insufficiency and the clinical diversity of rare diseases make it hard for deep learning models to be trained. Mortality prediction for these patients with different diseases can be viewed as a multi-task learning problem with insufficient data but a large number of tasks. On the other hand, insufficient training data makes it difficult to train task-specific modules in multi-task learning models. To address the challenges of data insufficiency and task diversity, we propose an initialization-sharing multi-task learning method (Ada-SiT). Ada-Sit can learn the parameter initialization and dynamically measure the tasks' similarities, used for fast adaptation. We use Ada-SiT to train long short-term memory networks (LSTM) based prediction models on longitudinal EHR data. The experimental results demonstrate that the proposed model is effective for mortality prediction of diverse rare diseases.

Cohort discovery and risk stratification for Alzheimer's disease: an electronic health record-based approach

BACKGROUND

We sought to leverage data routinely collected in electronic health records (EHRs), with the goal of developing patient risk stratification tools for predicting risk of developing Alzheimer's disease (AD).

METHOD

Using EHR data from the University of Michigan (UM) hospitals and consensus-based diagnoses from the Michigan Alzheimer's Disease Research Center, we developed and validated a cohort discovery tool for identifying patients with AD. Applied to all UM patients, these labels were used to train an EHR-based machine learning model for predicting AD onset within 10 years.

RESULTS

Applied to a test cohort of 1697 UM patients, the model achieved an area under the receiver operating characteristics curve of 0.70 (95% confidence interval = 0.63-0.77). Important predictive factors included cardiovascular factors and laboratory blood testing.

CONCLUSION

Routinely collected EHR data can be used to predict AD onset with modest accuracy. Mining routinely collected data could shed light on early indicators of AD appearance and progression.