PULSE learns representations from physiological time series.
problem Lack of effective pretraining objectives for physiological time series.
method Proposes a pretraining framework exploiting dynamical systems generative model.
result PULSE learns representations that distinguish semantic classes and improve transfer learning.
We developed a new approach for the analysis of physiological time series. An iterative convolution filter is used to decompose the time series into various components. Statistics of these components are extracted as features to characterize the mechanisms underlying the time series. Motivated by the studies that show …
Late fusion of clinical notes and physiological data improves ICU mortality prediction.
problem Improving ICU mortality prediction using multimodal data.
method Late fusion of clinical notes and physiological time series data with a deep learning architecture.
result Late fusion approach provides statistically significant improvement in mortality prediction performance.
PerCDL learns personalized dictionaries for physiological signals combining global and local structures.
problem Representing datasets with both global and local structures in human physiological signals.
method Personalized Convolutional Dictionary Learning (PerCDL) that combines a global and personalized local dictionary.
result PerCDL effectively learns interpretable representations for human locomotion data.
Personalized time series models improve drug response prediction.
problem Fitting time series models to cohorts ignoring individual variation.
method Multi-task learning applied to time series data with and without control inputs.
result Improved predictive accuracy and uncertainty estimation.
PHASE predicts surgical complications from physiological signals.
problem Predicting adverse surgical outcomes from physiological signals.
method Self-supervised transfer learning for physiological signals.
result PHASE outperforms other approaches in predicting five surgical complications.
DecompKAN improves time series forecasting accuracy and transparency.
problem Accurate and transparent time series forecasting in scientific domains.
method Combines decomposition, patching, normalization, and B-spline KAN edge functions.
result Achieves best or tied-best MSE on 20 of 36 comparisons across 9 datasets.
A new model uses GPs and latent force models to predict patient responses to drugs.
problem Challenges in modeling short-term effects of drugs on patient physiology.
method Hybrid Gaussian process with latent force model for joint modeling of patient physiology and drug effects.
result The model accurately predicts patient responses to three common drugs, showing competitive performance.
Support spinor machine extends SVM to handle spinor fields in time series data.
problem Handling nonstationary and nonlinear time series data for classification.
method Using wedge product to extend vector fields to spinor fields, extending SVM to support spinor machine.
result Support spinor machine outperforms SVM in one class classification of physiological time series data.
Paper uses topological data analysis for time series classification.
problem Classifying univariate time series data, especially physiological signals.
method Persistent homology for feature engineering, followed by machine learning.
result Higher accuracy achieved with fewer features compared to traditional methods.
Detects physiological patterns to hemodynamic stress using unsupervised deep learning.
problem Identify and characterize physiological responses to hemorrhage in raw vital sign data.
method Transform vital sign time series into latent space using unsupervised deep learning, identify clusters, and evaluate latent embeddings.
result Clusters in latent embeddings correspond to physiological response patterns matching physicians' intuition.
Here we propose a method, based on detrended covariance which we call detrended cross-correlation analysis (DXA), to investigate power-law cross-correlations between different simultaneously-recorded time series in the presence of non-stationarity. We illustrate the method by selected examples from physics, physiology,…
Deep learning model predicts ICU mortality by attending to missing data.
problem Handling irregular and missing data in ICU physiological time-series.
method Long Short-Term Memory architecture with attention mechanisms.
result Competitive and interpretable results on PhysioNet 2012 dataset.
End-to-end classifier predicts sepsis from clinical data.
problem Predicting sepsis onset from clinical data with high accuracy.
method Multitask Gaussian process RNN classifier for multivariate time series.
result Outperforms baselines, improving AUC and PRC.
U-Time uses a fully convolutional network for sleep stage classification.
problem Challenges in tuning and optimizing recurrent neural networks for sleep data.
method U-Time is a fully feed-forward deep learning approach based on U-Net architecture.
result U-Time outperforms state-of-the-art models for sleep stage classification.
Improved predictive models for AKI using intraoperative data.
problem Limited generalizability of existing AKI risk score models and lack of intraoperative data utilization.
method Machine learning and statistical analysis techniques, incorporating intraoperative physiologic time series data.
result The proposed model improves AKI risk prediction, achieving better AUROC and NRI.
New deep learning model for sparse, irregularly sampled multivariate time series.
problem Supervised learning with sparse and irregularly sampled multivariate time series.
method Interpolation-Prediction Networks: semi-parametric interpolation followed by a prediction network.
result Interpolation-Prediction Networks outperform baseline models in classification and regression tasks.
Study builds models to predict post-cardiac arrest outcomes using patient data.
problem Lack of accurate prognostication methods for patients resuscitated from cardiac arrest.
method Integrated electronic health records (EHR) and physiological time series (PTS) data to train machine learning classifiers.
result Combined EHR-PTS24 models outperformed models using either EHR or PTS24 alone in predicting survival and neurological outcomes.
Bayesian deep learning predicts emotion from heartbeat data.
problem Uncertainty in emotion prediction from physiological data.
method End-to-end deep learning model with Bayesian uncertainty estimation.
result Peak classification accuracy of 90% on benchmark datasets.
This paper proposes a nonparametric Bayesian method for exploratory data analysis and feature construction in continuous time series. Our method focuses on understanding shared features in a set of time series that exhibit significant individual variability. Our method builds on the framework of latent Diricihlet alloc…
Predicts individual septic shock children's vasoactive response using RNN.
problem Personalized physiologic responses to vasoactive titrations in septic shock children.
method Retrospective analysis of EMR data using a Recurrent Neural Network (RNN).
result RNN model predicted physiologic responses more accurately than a linear model.
SigTime learns interpretable signatures from time series data.
problem Discovering meaningful patterns in time series data with high complexity and limited interpretability.
method Jointly trains two Transformer models using shapelet-based and feature engineering representations.
result Learned shapelets serve as interpretable signatures for time series classification.
Bayesian model predicts emotion from fitness tracker heartbeat data.
problem Predicting emotional valence from consumer fitness tracker heartbeat data.
method End-to-end Bayesian deep learning model using PPG data.
result Peak F1 score of 0.7 for emotional valence classification.
New method detects change points in quasi-periodic signals without supervision.
problem Detecting change points in complex, non-harmonic signals.
method Optimal transport theory, topological analysis, bootstrap procedure.
result Successfully detects abnormal cardiac cycles in various arrhythmias.
New deep learning model tackles sparse clinical data.
problem Sparse and irregularly sampled multivariate time series in EHRs.
method Semi-parametric interpolation network followed by prediction network.
result Improved mortality and length of stay predictions.
Paper investigates personalization in emotion recognition from physiological data.
problem Emotion recognition from physiological signals.
method Features are extracted and ranked based on their effect on classification accuracy. Different classifiers are compared. Inter-subject variability and personalization effect are investigated through trial-based and subject-based cross-validation. A personalized model is introduced.
result Personalized model enhances emotional state prediction.
Self-supervised learning improves representation from EEG signals without labels.
problem Limited supervised data for EEG signal analysis.
method Predicting temporal context from unlabeled EEG time series.
result Self-supervised approach outperforms supervised methods in low data regimes.
The study creates benchmarks for clinical time series data to evaluate machine learning models.
problem Lack of publicly available benchmark data sets for healthcare research.
method Proposed four clinical prediction benchmarks using MIMIC-III data, evaluated various deep supervision and multitask training methods.
result Demonstrated the effectiveness of deep supervision, multitask training, and data-specific architectural modifications on neural models.
The study predicts acute hypotensive episodes using unsupervised learning and hashing.
problem Early detection of acute hypotensive episodes in ICU.
method Unsupervised representation learning and stratified locality sensitive hashing applied to multivariate time-series data.
result The method accurately predicts upcoming acute hypotensive episodes.
TQA improves prediction intervals for time series data by adjusting quantiles for both cross-sectional and longitudinal coverage.
problem Constructing reliable prediction intervals for cross-sectional time series data.
method Temporal Quantile Adjustment (TQA) method that adjusts the quantile in Conformal Prediction to account for both cross-sectional and longitudinal coverage.
result TQA improves longitudinal coverage while preserving cross-sectional coverage, as validated through extensive experimentation.
Model improves emotion recognition using multiple physiological signals.
problem Single physiological signal is insufficient for accurate emotion recognition.
method Fused multiple modal physiological signals (EEG, EMG, EOG) for emotion classification.
result Best classification accuracy of 94.42% on arousal and 94.02% on valence in two-class tasks.
Study builds complex network from multimodal physiological data.
problem Understanding dynamic interactions in biological systems.
method Network-based multimodal data fusion using recurrence plots and temporal metrics.
result Model accurately characterizes emotional states through physiological responses.
Hybrid Amortized Inference improves PPG model interpretability.
problem Tension between PPG biomarker accuracy and clinical interpretability.
method Introduces PPGen for biophysical PPG signal-physiological parameter relation, and HAI for fast, robust estimation.
result Hybrid Amortized Inference accurately infers physiological parameters from PPG signals.
Adversarial transfer learning improves stress assessment across users.
problem Transfer learning challenges in physiological biosignals.
method Disentangled nuisance-robust representations using adversarial networks.
result Adversarial framework enhances cross-subjects stress assessment.
We study the dynamics of correlation and variance in systems under the load of environmental factors. A universal effect in ensembles of similar systems under the load of similar factors is described: in crisis, typically, even before obvious symptoms of crisis appear, correlation increases, and, at the same time, vari…
Semi-supervised learning method augments minority class examples for robust anomaly detection in clinical signals.
problem Class imbalance in minority class instances impairs robustness of clinical analytics solutions.
method Intelligent augmentation of minority class examples to balance class distribution and construct a smooth decision boundary.
result The proposed method outperforms state-of-the-art algorithms in anomaly detection for clinical signals.
CLOPS improves deep learning for continuous physiological data.
problem Deep learning struggles with streaming, multi-sensor clinical data.
method CLOPS is a replay-based continual learning strategy.
result CLOPS outperforms state-of-the-art methods in continual learning scenarios.
Deep learning models predict postoperative complications more accurately than random forests.
problem Predicting postoperative complications to inform patient care decisions.
method Multi-task deep neural networks integrating intraoperative physiological data.
result Deep learning models improved prediction accuracy and provided interpretable risk factors.
Study classifies human stress using EEG, GSR, and PPG signals.
problem Classifying perceived human stress using physiological signals.
method Data acquisition, feature extraction (time domain), classification using multiple classifiers (SVM, Naive Bayes, MLP).
result Best classification accuracy of 75% achieved by MLP classifier.
Proposes a framework for extracting consistent physiological features across users.
problem Variability of biosignals across different users and tasks.
method Adversarial feature extractor for disentangled universal representations.
result Up to 8.8% improvement in average accuracy of classification.
The paper proposes a machine learning method to detect drivers' affective states using physiological signals.
problem Detecting and assessing drivers' affective states to improve driving safety and well-being.
method Multiview multi-task machine learning approach using physiological signals.
result Accounting for drive-specific differences significantly improves model performance.
UNHaP removes noise from physiological events using Hawkes processes.
problem Challenges in identifying true events from spurious ones in physiological signal analysis.
method UNHaP uses marked Hawkes processes to distinguish and unmix true events from noise.
result UNHaP significantly reduces false detection rates and enhances event understanding.
Neural recordings are nonstationary time series, i.e. their properties typically change over time. Identifying specific changes, e.g. those induced by a learning task, can shed light on the underlying neural processes. However, such changes of interest are often masked by strong unrelated changes, which can be of physi…
This study benchmarks changepoint detection algorithms on cardiac time series data.
problem Identifying state changes in cardiac time series for disease classification.
method Comparison of 8 changepoint detection algorithms on artificial and real cardiac time series data.
result RMDM algorithm achieved highest true positive rate and cross validated accuracy for classification.
Neurons predict future scalar inputs by learning top modes of lag vectors.
problem Predicting future scalar inputs with physiological delays.
method Normal Mode Decomposition to extract independently evolving modes.
result Temporal filters of neurons correspond to left eigenvectors of a generalized eigenvalue problem.
Bayesian model detects internal bleeding in ICU patients.
problem Late detection of internal bleeding in ICU patients.
method Bayesian regime switching model analyzing vitals and labs.
result Probabilistic assessment of patient's physiological state.
PSEUDo learns patterns in multivariate time series with locality-sensitive hashing and relevance feedback.
problem Efficient pattern detection in large, multi-track sequential data with high variance and lack of ground truth.
method Query-aware locality-sensitive hashing for feature learning, sub-linear training and inference time.
result PSEUDo achieves sub-linear time efficiency for pattern modeling and comparison of 10,000 multivariate time series.
Predicts physiologically acceptable states for pediatric ICU discharge.
problem Determining physiologically stable states for pediatric ICU discharge.
method Computed PASS values from hr, sbp, and dbp measurements, compared to age-normal and polynomial regression predictions.
result RNN model predictions were more accurate than age-normal vitals.