Handling missing data in time series classification remains a significant challenge in various domains. Traditional methods often rely on imputation, which may introduce bias or fail to capture the underlying temporal dynamics. In this paper, we propose TANDEM (Temporal Attention-guided Neural Differential Equations for Missingness), an attention-guided neural differential equation framework that effectively classifies time series data with missing values. Our approach integrates raw observation, interpolated control path, and continuous latent dynamics through a novel attention mechanism, allowing the model to focus on the most informative aspects of the data. We evaluate TANDEM on 30 benchmark datasets and a real-world medical dataset, demonstrating its superiority over existing state-of-the-art methods. Our framework not only improves classification accuracy but also provides insights into the handling of missing data, making it a valuable tool in practice.

Time series modeling and analysis have become critical in various domains. Conventional methods such as RNNs and Transformers, while effective for discrete-time and regularly sampled data, face significant challenges in capturing the continuous dynamics and irregular sampling patterns inherent in real-world scenarios. Neural Differential Equations (NDEs) represent a paradigm shift by combining the flexibility of neural networks with the mathematical rigor of differential equations. This paper presents a comprehensive review of NDE-based methods for time series analysis, including neural ordinary differential equations, neural controlled differential equations, and neural stochastic differential equations. We provide a detailed discussion of their mathematical formulations, numerical methods, and applications, highlighting their ability to model continuous-time dynamics. Furthermore, we address key challenges and future research directions. This survey serves as a foundation for researchers and practitioners seeking to leverage NDEs for advanced time series analysis.

The development and deployment of machine learning (ML) models for biomedical research and healthcare currently lacks standard methodologies. Although tools for model replication are numerous, without a unifying blueprint it remains difficult to scientifically reproduce predictive ML models for any number of reasons (e.g., assumptions regarding data distributions and preprocessing, unclear test metrics, etc.) and ultimately, questions around generalizability and transportability are not readily answered. To facilitate scientific reproducibility, we built upon the Predictive Model Markup Language (PMML) to capture essential information. As a key component of the PREdictive Model Index and Exchange REpository (PREMIERE) platform, we present the Automated Metadata Pipeline (AMP) for conversion of a given predictive ML model into an extended PMML file that autocompletes an ML-based checklist, assessing model elements for interoperability and reproducibility. We demonstrate this pipeline on multiple test cases with three different ML algorithms and health-related datasets, providing a foundation for future predictive model reproducibility, sharing, and comparison. Problem: The development and deployment of machine learning (ML) models for biomedical research and healthcare currently lack standard methodologies, leading to problems of scientific reproducibility and interoperability. What is Already Known: Although there are many tools for model replication, without a unifying blueprint, it remains difficult to scientifically reproduce predictive ML models for any number of reasons. Moreover, questions around generalizability and transportability are not readily answered. What this Paper Adds: This study builds upon the Predictive Model Markup Language (PMML) to capture essential information and presents the Automated Metadata Pipeline (AMP) for conversion of a given predictive ML model into an extended PMML file that auto-completes an ML-based checklist, assessing model elements for interoperability and reproducibility. We demonstrate this pipeline on multiple test cases with three different ML algorithms and health-related datasets. The proposed AMP provides a framework for automating the completion and evaluation of a comprehensive ML model checklist, increasing compliance and ultimately, predictive model reproducibility, sharing, and comparison of predictive models by ensuring all appropriate information is available.

Digital health data from consumer wearable devices and smartphones have the potential to improve our understanding of mental illness. However, in conditions like depression, there is not yet a consistent uniform measurement tool whose result can be reliably used as a gold standard measure of depression severity. This work seeks to specify what symptoms and dimensions of depression can be detected using vitals, activity, and sleep monitored by consumer wearable devices. Machine learning models are fit to digital health data and used to detect responses to individual questions from surveys (self-reports) as well as summary scores from these self-reports. For high performing models, feature importance is investigated. Analysis is conducted on preliminary data from 99 participants of an ongoing study with data from the Apple Watch and iPhone along with validated self-reports relevant to depression severity, anhedonia severity, and sleep quality. Receiver operator characteristic area under the curve (ROC AUC) and average precision are used to assess model performance. The digital health sensor data investigated was found to significantly detect five of 74 measures, including overall depression severity and specific symptoms like poor appetite, aspects of anhedonia, and sleep timings (ROC AUC between 0.63 and 0.72). The features these models use in detection vary per detection task and suggest further areas for investigation to specify the right features to look at per symptom.

Early detection of heart failure (HF) can provide patients with the opportunity for more timely intervention and better disease management, as well as efficient use of healthcare resources. Recent machine learning (ML) methods have shown promising performance on diagnostic prediction using temporal sequences from electronic health records (EHRs). In practice, however, these models may not generalize to other populations due to dataset shift. Shifts in datasets can be attributed to a range of factors such as variations in demographics, data management methods, and healthcare delivery patterns. In this paper, we use unsupervised adversarial domain adaptation methods to adaptively reduce the impact of dataset shift on cross-institutional transfer performance. The proposed framework is validated on a next-visit HF onset prediction task using a BERT-style Transformer-based language model pre-trained with a masked language modeling (MLM) task. Our model empirically demonstrates superior prediction performance relative to non-adversarial baselines in both transfer directions on two different clinical event sequence data sources.

To better understand the challenges of generally implementing and adapting computational phenotyping approaches, the performance of a Phenotype KnowledgeBase (PheKB) algorithm for rheumatoid arthritis (RA) was evaluated on a University of California, Los Angeles (UCLA) patient population, focusing on examining its performance on ambiguous cases. The algorithm was evaluated on a cohort of 4,766 patients, along with a chart review of 300 patients by rheumatologists against accepted diagnostic guidelines. The performance revealed low sensitivity towards specific subtypes of positive RA cases, which suggests revisions in features used for phenotyping. A close examination of select cases also indicated a significant portion of patients with missing data, drawing attention to the need to consider data integrity as an integral part of phenotyping pipelines, as well as issues around the usability of various codes for distinguishing cases. We use patterns in the PheKB algorithm's errors to further demonstrate important considerations when designing a phenotyping algorithm.

Asthma is the most prevalent chronic disease among pediatrics, as it is the leading cause of student absenteeism and hospitalization for those under the age of 15. To address the significant need to manage this disease in children, the authors present a mobile health (mHealth) system that determines the risk of an asthma attack through physiological and environmental wireless sensors and representational state transfer application program interfaces (RESTful APIs). The data is sent from wireless sensors to a smartwatch application (app) via a Health Insurance Portability and Accountability Act (HIPAA) compliant cryptography framework, which then sends data to a cloud for real-time analytics. The asthma risk is then sent to the smartwatch and provided to the user via simple graphics for easy interpretation by children. After testing the safety and feasibility of the system in an adult with moderate asthma prior to testing in children, it was found that the analytics model is able to determine the overall asthma risk (high, medium, or low risk) with an accuracy of 80.10±14.13%. Furthermore, the features most important for assessing the risk of an asthma attack were multifaceted, highlighting the importance of continuously monitoring different wireless sensors and RESTful APIs. Future testing this asthma attack risk prediction system in pediatric asthma individuals may lead to an effective self-management asthma program.

Given the increasing emphasis on delivering high-quality, cost-efficient healthcare, improved methodologies are needed to measure the accuracy and utility of ordered diagnostic examinations in achieving the appropriate diagnosis. Here, we present a data-driven approach for performing automated quality assessment of radiologic interpretations using other clinical information (e.g., pathology) as a reference standard for individual radiologists, subspecialty sections, imaging modalities, and entire departments. Downstream diagnostic conclusions from the electronic medical record are utilized as "truth" to which upstream diagnoses generated by radiology are compared. The described system automatically extracts and compares patient medical data to characterize concordance between clinical sources. Initial results are presented in the context of breast imaging, matching 18 101 radiologic interpretations with 301 pathology diagnoses and achieving a precision and recall of 84% and 92%, respectively. The presented data-driven method highlights the challenges of integrating multiple data sources and the application of information extraction tools to facilitate healthcare quality improvement.

Patient portals have the potential to provide content that is specifically tailored to a patient's information needs based on diagnoses and other factors. In this work, we conducted a survey of 41 lung cancer patients at an outpatient lung cancer clinic at the medical center of the University of California Los Angeles, to gain insight into these perceived information needs and opinions on the design of a portal to fulfill them. We found that patients requested access to information related to diagnosis and imaging, with more than half of the patients reporting that they did not anticipate an increase in anxiety due to access to medical record information via a portal. We also found that patient educational background did not lead to a significant difference in desires for explanations of reports and definitions of terms.

OBJECTIVE: With the increased routine use of advanced imaging in clinical diagnosis and treatment, it has become imperative to provide patients with a means to view and understand their imaging studies. We illustrate the feasibility of a patient portal that automatically structures and integrates radiology reports with corresponding imaging studies according to several information orientations tailored for the layperson. METHODS: The imaging patient portal is composed of an image processing module for the creation of a timeline that illustrates the progression of disease, a natural language processing module to extract salient concepts from radiology reports (73% accuracy, F1 score of 0.67), and an interactive user interface navigable by an imaging findings list. The portal was developed as a Java-based web application and is demonstrated for patients with brain cancer. RESULTS AND DISCUSSION: The system was exhibited at an international radiology conference to solicit feedback from a diverse group of healthcare professionals. There was wide support for educating patients about their imaging studies, and an appreciation for the informatics tools used to simplify images and reports for consumer interpretation. Primary concerns included the possibility of patients misunderstanding their results, as well as worries regarding accidental improper disclosure of medical information. CONCLUSIONS: Radiologic imaging composes a significant amount of the evidence used to make diagnostic and treatment decisions, yet there are few tools for explaining this information to patients. The proposed radiology patient portal provides a framework for organizing radiologic results into several information orientations to support patient education.

Imaging has become a prevalent tool in the diagnosis and treatment of many diseases, providing a unique in vivo, multi-scale view of anatomic and physiologic processes. With the increased use of imaging and its progressive technical advances, the role of imaging informatics is now evolving--from one of managing images, to one of integrating the full scope of clinical information needed to contextualize and link observations across phenotypic and genotypic scales. Several challenges exist for imaging informatics, including the need for methods to transform clinical imaging studies and associated data into structured information that can be organized and analyzed. We examine some of these challenges in establishing imaging-based observational databases that can support the creation of comprehensive disease models. The development of these databases and ensuing models can aid in medical decision making and knowledge discovery and ultimately, transform the use of imaging to support individually-tailored patient care.

We believe that by adapting architectures to fit the requirements of a given application domain, we can significantly improve the efficiency of computation. To validate the idea for our application domain, we evaluate a wide spectrum of commodity computing platforms to quantify the potential benefits of heterogeneity and customization for the domain-specific applications. In particular, we choose medical imaging as the application domain for investigation, and study the application performance and energy efficiency across a diverse set of commodity hardware platforms, such as general-purpose multi-core CPUs, massive parallel many-core GPUs, low-power mobile CPUs and fine-grain customizable FPGAs. This study leads to a number of interesting observations that can be used to guide further development of domain-specific architectures.

Time series subsequence matching (or signal searching) has importance in a variety of areas in health care informatics. These areas include case-based diagnosis and treatment as well as the discovery of trends and correlations between data. Much of the traditional research in signal searching has focused on high dimensional R-NN matching. However, the results of R-NN are often small and yield minimal information gain; especially with higher dimensional data. This paper proposes a randomized Monte Carlo sampling method to broaden search criteria such that the query results are an accurate sampling of the complete result set. The proposed method is shown both theoretically and empirically to improve information gain. The number of query results are increased by several orders of magnitude over approximate exact matching schemes and fall within a Gaussian distribution. The proposed method also shows excellent performance as the majority of overhead added by sampling can be mitigated through parallelization. Experiments are run on both simulated and real-world biomedical datasets.

Due to the increasingly data-intensive clinical environment, physicians now have unprecedented access to detailed clinical information from a multitude of sources. However, applying this information to guide medical decisions for a specific patient case remains challenging. One issue is related to presenting information to the practitioner: displaying a large (irrelevant) amount of information often leads to information overload. Next-generation interfaces for the electronic health record (EHR) should not only make patient data easily searchable and accessible, but also synthesize fragments of evidence documented in the entire record to understand the etiology of a disease and its clinical manifestation in individual patients. In this paper, we describe our efforts toward creating a context-based EHR, which employs biomedical ontologies and (graphical) disease models as sources of domain knowledge to identify relevant parts of the record to display. We hypothesize that knowledge (e.g., variables, relationships) from these sources can be used to standardize, annotate, and contextualize information from the patient record, improving access to relevant parts of the record and informing medical decision making. To achieve this goal, we describe a framework that aggregates and extracts findings and attributes from free-text clinical reports, maps findings to concepts in available knowledge sources, and generates a tailored presentation of the record based on the information needs of the user. We have implemented this framework in a system called Adaptive EHR, demonstrating its capabilities to present and synthesize information from neurooncology patients. This paper highlights the challenges and potential applications of leveraging disease models to improve the access, integration, and interpretation of clinical patient data.

The Clinical Outcomes Assessment Toolkit (COAT) was created through a collaboration between the University of California, Los Angeles and Brigham and Women's Hospital to address the challenge of gathering, formatting, and abstracting data for clinical outcomes and performance measurement research. COAT provides a framework for the development of information pipelines to transform clinical data from its original structured, semi-structured, and unstructured forms to a standardized format amenable to statistical analysis. This system includes a collection of clinical data structures, reusable utilities for information analysis and transformation, and a graphical user interface through which pipelines can be controlled and their results audited by nontechnical users. The COAT architecture is presented, as well as two case studies of current implementations in the domain of prostate cancer outcomes assessment.

OBJECTIVES: The College of American Pathologists (CAP) Category 1 quality measures, tumor stage, Gleason score, and surgical margin status, are used by physicians and cancer registrars to categorize patients into groups for clinical trials and treatment planning. This study was conducted to evaluate the effectiveness of an application designed to automatically extract these quality measures from the postoperative pathology reports of patients having undergone prostatectomies for treatment of prostate cancer. DESIGN: An application was developed with the Clinical Outcomes Assessment Toolkit that uses an information pipeline of regular expressions and support vector machines to extract CAP Category 1 quality measures. System performance was evaluated against a gold standard of 676 pathology reports from the University of California at Los Angeles Medical Center and Brigham and Women's Hospital. To evaluate the feasibility of clinical implementation, all pathology reports were gathered using administrative codes with no manual preprocessing of the data performed. MEASUREMENTS: The sensitivity, specificity, and overall accuracy of system performance were measured for all three quality measures. Performance at both hospitals was compared, and a detailed failure analysis was conducted to identify errors caused by poor data quality versus system shortcomings. RESULTS: Accuracies for Gleason score were 99.7%, tumor stage 99.1%, and margin status 97.2%, for an overall accuracy of 98.67%. System performance on data from both hospitals was comparable. Poor clinical data quality led to a decrease in overall accuracy of only 0.3% but accounted for 25.9% of the total errors. CONCLUSION: Despite differences in document format and pathologists' reporting styles, strong system performance indicates the potential of using a combination of regular expressions and support vector machines to automatically extract CAP Category 1 quality measures from postoperative prostate cancer pathology reports.

An increasing amount of data is now accrued in medical information systems; however, the organization of this data is still primarily driven by data source, and does not support the cognitive processes of physicians. As such, new methods to visualize patient medical records are becoming imperative in order to assist physicians with clinical tasks and medical decision-making. The TimeLine system is a problem-centric temporal visualization for medical data: information contained with medical records is reorganized around medical disease entities and conditions. Automatic construction of the TimeLine display from existing clinical repositories occurs in three steps: 1) data access, which uses an eXtensible Markup Language (XML) data representation to handle distributed, heterogeneous medical databases; 2) data mapping and reorganization, reformulating data into hierarchical, problemcentric views; and 3) data visualization, which renders the display to a target presentation platform. Leveraging past work, we describe the latter two components of the TimeLine system in this paper, and the issues surrounding the creation of medical problems lists and temporal visualization of medical data. A driving factor in the development of TimeLine was creating a foundation upon which new data types and the visualization metaphors could be readily incorporated.

The development of comprehensive picture archive and communication systems (PACS) has mainly been limited to proprietary developments by vendors, though a number of freely available software projects have addressed specific image management tasks. The openSourcePACS project aims to provide an open source, common foundation upon which not only can a basic PACS be readily implemented, but to also support the evolution of new PACS functionality through the development of novel imaging applications and services. openSourcePACS consists of four main software modules: 1) image order entry, which enables the ordering and tracking of structured image requisitions; 2) an agent-based image server framework that coordinates distributed image services including routing, image processing, and querying beyond the present digital image and communications in medicine (DICOM) capabilities; 3) an image viewer, supporting standard display and image manipulation tools, DICOM presentation states, and structured reporting; and 4) reporting and result dissemination, supplying web-based widgets for creating integrated reports. All components are implemented using Java to encourage cross-platform deployment. To demonstrate the usage of openSourcePACS, a preliminary application supporting primary care/specialist communication was developed and is described herein. Ultimately, the goal of openSourcePACS is to promote the wide-scale development and usage of PACS and imaging applications within academic and research communities.

This paper presents an incremental diagnosis method (IDM) to detect a medical condition with the minimum wearable sensor usage by dynamically adjusting the sensor set based on the patient's state in his/her natural environment. The IDM, comprised of a naive Bayes classifier generated by supervised training with Gaussian clustering, is developed to classify patient motion in-context (due to a medical condition) and in real-time using a wearable sensor system. The IDM also incorporates a utility function, which is a simple form of expert knowledge and user preferences in sensor selection. Upon initial in-context detection, the utility function decides which sensor is to be activated next. High-resolution in-context detection with minimum sensor usage is possible because the necessary sensor can be activated or requested at the appropriate time. As a case study, the IDM is demonstrated in detecting different severity levels of a limp with minimum usage of high diagnostic resolution sensors.

Prefetching methods have traditionally been used to restore archived images from picture archiving and communication systems to diagnostic imaging workstations prior to anticipated need, facilitating timely comparison of historical studies and patient management. The authors describe a problem-oriented prefetching scheme, detailing 1) a mechanism supporting selection of patients for prefetching via characterizations of clinical problems, using multiple data sources (picture archiving and communication systems, hospital information systems, and radiology information systems), classifying patients into cohorts on the basis of their medical conditions (e.g., lung cancer); and 2) prefetching of multimedia data (imaging, laboratory, and medical reports) from clinical databases to enable the viewing of an integrated patient record. Preliminary evaluation of the prefetching algorithm using classic information retrieval measures showed that the system had high recall (100 percent), correctly identifying and retrieving data for all patients belonging to a target cohort, but low precision (50 percent). A key finding during testing was that the recall of the system was increased through the use of multiple data sources (compared with one data source), because of better patient descriptors. Medical problems and patient cohorts were more specifically defined by combining information from heterogeneous databases.