Peripheral Artery Disease (PAD) is a common atherosclerotic condition that is underdiagnosed due to the lack of accessible screening options. Photoplethysmography (PPG) serves as a potentially valuable tool in accessible screening for PAD due to its ubiquitous nature and ability to be measured on a smartphone. However, the relationship between PPG and PAD is underexplored. In this paper, we seek to identify features of a PPG signal that correlate with PAD. In an analysis of 5,237 legs from$\mathrm{N}=2,362$unique patients, we find significant correlations with multiple different features and the ankle-brachial index (ABI), which is used to diagnose PAD. Additionally, these features agree with physiological explanations of PAD and how the disease affects blood flow. These results set up the ability of future work to develop an accessible screening tool for PAD that uses physiologically relevant features of PPG morphology.

OBJECTIVE: Unplanned readmissions following a hospitalization remain common despite significant efforts to curtail these. Wearable devices may offer help identify patients at high risk for an unplanned readmission. MATERIALS AND METHODS: We conducted a multi-center retrospective cohort study using data from the All of Us data repository. We included subjects with wearable data and developed a baseline Feedforward Neural Network (FNN) model and a Long Short-Term Memory (LSTM) time-series deep learning model to predict daily, unplanned rehospitalizations up to 90 days from discharge. In addition to demographic and laboratory data from subjects, post-discharge data input features include wearable data and multiscale entropy features based on intraday wearable time series. The most significant features in the LSTM model were determined by permutation feature importance testing. RESULTS: In sum, 612 patients met inclusion criteria. The complete LSTM model had a higher area under the receiver operating characteristic curve than the FNN model (0.83 vs 0.795). The 5 most important input features included variables from multiscale entropy (steps) and number of active steps per day. DISCUSSION: Data available from wearable devices can improve ability to predict readmissions. Prior work has focused on predictors available up to discharge or on additional data abstracted from wearable devices. Our results from 35 institutions highlight how multiscale entropy can improve readmission prediction and may impact future work in this domain. CONCLUSION: Wearable data and multiscale entropy can improve prediction of a deep-learning model to predict unplanned 90-day readmissions. Prospective studies are needed to validate these findings.

OBJECTIVE: Sepsis has a high rate of 30-day unplanned readmissions. Predictive modeling has been suggested as a tool to identify high-risk patients. However, existing sepsis readmission models have low predictive value and most predictive factors in such models are not actionable. MATERIALS AND METHODS: Data from patients enrolled in the AllofUs Research Program cohort from 35 hospitals were used to develop a multicenter validated sepsis-related unplanned readmission model that incorporates clinical and social determinants of health (SDH) to predict 30-day unplanned readmissions. Sepsis cases were identified using concepts represented in the Observational Medical Outcomes Partnership. The dataset included over 60 clinical/laboratory features and over 100 SDH features. RESULTS: Incorporation of SDH factors into our model of clinical and demographic features improves model area under the receiver operating characteristic curve (AUC) significantly (from 0.75 to 0.80; P < .001). Model-agnostic interpretability techniques revealed demographics, economic stability, and delay in getting medical care as important SDH predictive features of unplanned hospital readmissions. DISCUSSION: This work represents one of the largest studies of sepsis readmissions using objective clinical data to date (8935 septic index encounters). SDH are important to determine which sepsis patients are more likely to have an unplanned 30-day readmission. The AllofUS dataset provides granular data from a diverse set of individuals, making this model potentially more generalizable than prior models. CONCLUSION: Use of SDH improves predictive performance of a model to identify which sepsis patients are at high risk of an unplanned 30-day readmission.

Major depressive disorder is a common psychiatric illness. At present, there are no objective, non-verbal, automated markers that can reliably track treatment response. Here, we explore the use of video analysis of facial expressivity in a cohort of severely depressed patients before and after deep brain stimulation (DBS), an experimental treatment for depression. We introduced a set of variability measurements to obtain unsupervised features from muted video recordings, which were then leveraged to build predictive models to classify three levels of severity in the patients' recovery from depression. Multiscale entropy was utilized to estimate the variability in pixel intensity level at various time scales. A dynamic latent variable model was utilized to learn a low-dimensional representation of factors that describe the dynamic relationship between high-dimensional pixels in each video frame and over time. Finally, a novel elastic net ordinal regression model was trained to predict the severity of depression, as independently rated by standard rating scales. Our results suggest that unsupervised features extracted from these video recordings, when incorporated in an ordinal regression predictor, can discriminate different levels of depression severity during ongoing DBS treatment. Objective markers of patient response to treatment have the potential to standardize treatment protocols and enhance the design of future clinical trials.

Detection of atrial fibrillation (AF), a type of cardiac arrhythmia, is difficult since many cases of AF are usually clinically silent and undiagnosed. In particular paroxysmal AF is a form of AF that occurs occasionally, and has a higher probability of being undetected. In this work, we present an attention based deep learning framework for detection of paroxysmal AF episodes from a sequence of windows. Time-frequency representation of 30 seconds recording windows, over a 10 minute data segment, are fed sequentially into a deep convolutional neural network for image-based feature extraction, which are then presented to a bidirectional recurrent neural network with an attention layer for AF detection. To demonstrate the effectiveness of the proposed framework for transient AF detection, we use a database of 24 hour Holter Electrocardiogram (ECG) recordings acquired from 2850 patients at the University of Virginia heart station. The algorithm achieves an AUC of 0.94 on the testing set, which exceeds the performance of baseline models. We also demonstrate the cross-domain generalizablity of the approach by adapting the learned model parameters from one recording modality (ECG) to another (photoplethysmogram) with improved AF detection performance. The proposed high accuracy, low false alarm algorithm for detecting paroxysmal AF has potential applications in long-term monitoring using wearable sensors.

Physiological variables, such as heart rate (HR), blood pressure (BP) and respiration (RESP), are tightly regulated and coupled under healthy conditions, and a break-down in the coupling has been associated with aging and disease. We present an approach that incorporates physiological modeling within a switching linear dynamical systems (SLDS) framework to assess the various functional components of the autonomic regulation through transfer function analysis of nonstationary multivariate time series of vital signs. We validate our proposed SLDS-based transfer function analysis technique in automatically capturing 1) changes in baroreflex gain due to postural changes in a tilt-table study including ten subjects, and 2) the effect of aging on the autonomic control using HR/RESP recordings from 40 healthy adults. Next, using HR/BP time series of more than 450 adult ICU patients, we show that our technique can be used to reveal coupling changes associated with severe sepsis (AUC = 0.74, sensitivity = 0.74, specificity = 0.60). Our findings indicate that reduced HR/BP coupling is significantly associated with severe sepsis even after adjusting for clinical interventions (P 0.001). These results demonstrate the utility of our approach in phenotyping complex vital-sign dynamics, and in providing mechanistic hypotheses in terms of break-down of autoregulatory systems under healthy and disease conditions.

We focus on an important problem of predicting the so-called “patient flow” from longitudinal electronic health records (EHRs), which has not been explored via existing machine learning techniques. We develop a point process based framework for modeling patient flow through various care units (CUs) and jointly predicting patients' destination CUs and duration days. We propose a novel discriminative learning algorithm aiming at improving the prediction of transition events in the case of sparse data. By parameterizing the proposed model as mutually-correcting processes, we formulate the estimation problem via generalized linear models and solve it based on alternating direction method of multipliers (ADMM). We achieve simultaneous feature selection and learning by adding a group-lasso regularizer to the ADMM algorithm. Additionally, we synthesize auxiliary training data for the classes with extremely few samples, and improve the robustness of our learning method to the problem of data imbalance.

Over the past decade, the rate of care unit (CU) use in the United States has been increasing. With an aging population and ever-growing demand for medical care, effective management of patients' transitions among different care facilities will prove indispensible for shortening the length of hospital stays, improving patient outcomes, allocating critical care resources, and reducing preventable re-admissions. In this paper, we focus on an important problem of predicting the so-called “patient flow” from longitudinal electronic health records (EHRs), which has not been explored via existing machine learning techniques. By treating a sequence of transition events as a point process, we develop a novel framework for modeling patient flow through various CUs and jointly predicting patients' destination CUs and duration days. Instead of learning a generative point process model via maximum likelihood estimation, we propose a novel discriminative learning algorithm aiming at improving the prediction of transition events in the case of sparse data. By parameterizing the proposed model as a mutually-correcting process, we formulate the estimation problem via generalized linear models, which lends itself to efficient learning based on alternating direction method of multipliers (ADMM). Furthermore, we achieve simultaneous feature selection and learning by adding a group-lasso regularizer to the ADMM algorithm. Additionally, for suppressing the negative influence of data imbalance on the learning of model, we synthesize auxiliary training data for the classes with extremely few samples, and improve the robustness of our learning method accordingly. Testing on real-world data, we show that our method obtains superior performance in terms of accuracy of predicting the destination CU transition and duration of each CU occupancy.

Clinical data management systems typically provide caregiver teams with useful information, derived from large, sometimes highly heterogeneous, data sources that are often changing dynamically. Over the last decade there has been a significant surge in interest in using these data sources, from simply re-using the standard clinical databases for event prediction or decision support, to including dynamic and patient-specific information into clinical monitoring and prediction problems. However, in most cases, commercial clinical databases have been designed to document clinical activity for reporting, liability and billing reasons, rather than for developing new algorithms. With increasing excitement surrounding "secondary use of medical records" and "Big Data" analytics, it is important to understand the limitations of current databases and what needs to change in order to enter an era of "precision medicine." This review article covers many of the issues involved in the collection and preprocessing of critical care data. The three challenges in critical care are considered: compartmentalization, corruption, and complexity. A range of applications addressing these issues are covered, including the modernization of static acuity scoring; on-line patient tracking; personalized prediction and risk assessment; artifact detection; state estimation; and incorporation of multimodal data sources such as genomic and free text data.

Cardiovascular variables such as heart rate (HR) and blood pressure (BP) are regulated by an underlying control system, and therefore, the time series of these vital signs exhibit rich dynamical patterns of interaction in response to external perturbations (e.g., drug administration), as well as pathological states (e.g., onset of sepsis and hypotension). A question of interest is whether "similar" dynamical patterns can be identified across a heterogeneous patient cohort, and be used for prognosis of patients' health and progress. In this paper, we used a switching vector autoregressive framework to systematically learn and identify a collection of vital sign time series dynamics, which are possibly recurrent within the same patient and may be shared across the entire cohort. We show that these dynamical behaviors can be used to characterize the physiological "state" of a patient. We validate our technique using simulated time series of the cardiovascular system, and human recordings of HR and BP time series from an orthostatic stress study with known postural states. Using the HR and BP dynamics of an intensive care unit (ICU) cohort of over 450 patients from the MIMIC II database, we demonstrate that the discovered cardiovascular dynamics are significantly associated with hospital mortality (dynamic modes 3 and 9, p=0.001, p=0.006 from logistic regression after adjusting for the APACHE scores). Combining the dynamics of BP time series and SAPS-I or APACHE-III provided a more accurate assessment of patient survival/mortality in the hospital than using SAPS-I and APACHE-III alone (p=0.005 and p=0.045). Our results suggest that the discovered dynamics of vital sign time series may contain additional prognostic value beyond that of the baseline acuity measures, and can potentially be used as an independent predictor of outcomes in the ICU.

Robust navigation and mining of physiologic time series databases often requires finding similar temporal patterns of physiological responses. Detection of these complex physiological patterns not only enables demarcation of important clinical events but can also elucidate hidden dynamical structures that may be suggestive of disease processes. Some specific examples where this physiological signal search may be useful include real-time detection of cardiac arrhythmias, sleep staging or detection of seizure onset. In all these cases, being able to identify a cohort of patients who exhibit similar physiological dynamics could be useful in prognosis and informing treatment strategies. However, pattern recognition for physiological time series is complicated by changes between operating regimes and measurement artifacts. Here we briefly describe an approach we have developed for distributed identification of dynamical patterns in physiological time series using a switching linear dynamical system (SLDS). We present a fast and memory-efficient algorithm for learning and retrieval of phenotypic dynamics in large clinical time series databases. Through simulation we show that the proposed algorithm is at least an order of magnitude faster that the state of the art, and provide encouraging preliminary results based on real recordings of vital sign time series from the Multiparameter Intelligent Monitoring in Intensive Care (MIMIC-II) database.

The research Weather Surveillance Radar-1988 Doppler locally operated by the National Severe Storms Laboratory in Norman, OK, has the unique capability of collecting massive volumes of Level I time-series data over many hours, which provides a rich environment for evaluating our new postprocessing algorithms. In this letter, an approach of identifying tornado vortices in Doppler spectra is proposed and investigated using eigen analysis, cluster estimation, and fuzzy logic technique.

In this paper, we present a new particle filtering (PF) algorithm for visual target tracking where Galerkin's projection method is used to generate the proposal distribution. Galerkin's method is a numerical approach to approximate the solution of a partial differential equation (PDE). By leveraging this method in concert with L2theory and the FFT, we obtain a new proposal which directly approximates the true state posterior distribution and is fundamentally different from various local linearizations or Kalman filter-based proposals. We apply this improved PF algorithm to track a human head in a video sequence. As predicted by theory and demonstrated by our experimental results, this new algorithm is highly effective for tracking targets which exhibit complex kinematics. The new proposal distribution given here captures the high probability area in the state space, thereby gleaning increased support from the true posterior distribution.