Research
Bio-signal processing
I specialize in deep learning algorithms for physiological signals collected from wearable devices, including electrocardiograms (ECG), photoplethysmography (PPG), phonocardiograms (PCG, heart sounds), and arterial blood pressure (ABP). Many of these solutions have been productionized and ranked first place at the PhysioNet Challenges.
- 2021 – Interpretable Additive Recurrent Neural Networks For Multivariate Clinical Time Series
- 2020 – A Wide and Deep Transformer Neural Network for 12-Lead ECG Classification
- 2019 – A Multi-Task Imputation and Classification Neural Architecture for Early Prediction of Sepsis from Multivariate Clinical Time Series
- 2018 – Densely connected convolutional networks for detection of atrial fibrillation from short single-lead ECG recordings
- 2018 – Analyzing single-lead short ECG recordings using dense convolutional neural networks and feature-based post-processing to detect atrial fibrillation
- 2016 – Ensemble of feature-based and deep learning-based classifiers for detection of abnormal heart sounds
Interpretable machine learning
A key requirement of clinical models is that they have to be explainable so that the nurse or clinician can understand the risk factors. The models should also be editable so we can identify when then algorithm makes an error and fix it before deployment. I created the Interpretable-RNN, which is an additive sequence learning model that achieves both high accuracy and is fully glassbox.
I developed the neural decision tree model which is a neural representation of highly interpretable gradient boosted decision trees. This allowed us to scale an interpretable class of models in a federated privacy preserving architecture.
Disease prediction
I built realtime clinical risk algorithms that are deployed on patient monitors and central stations.
Early warning systems like the Hemodynamic Stability Index identifies ICU patients that will need fluids or pressors in the near future. The HSI model is designed for Philips bedside patient monitors and is the flagship machine learning model in hemodynamics analytics products. This work was extended with federated learning, a privacy preserving technique that was used to train the model across hundreds of ICUs.
Machine learning can also be used to improve clinical trial design by finding the right patient at the right time. I led a collaboration with Bayer Pharmaceuticals to optimize clinical trials investigating drugs to treat respiratory distress.
Patient similarity
Patient similarity aims to find cohorts based on vital signs, labs, medical history, and treatments for case-based reasoning. How can we learn useful similarity functions that compare physiological state between patients? We developed a similarity metric to find patients-like-me in electronic health records.
Prior medical knowledge is widely available in electronic health records but are not utilized because of practical constraints on data availability or cost of acquiring the data to make inferences. I developed a a novel approach to learn from metadata called prior knowledge-guided learning using a mixture-of-experts model where gating probabilities are tuned by a nearest neighbors graph adjacency matrix.
Neural engineering
My research in neural engineering spans computational modeling, animal models, and the analysis of brain signals like evoked potentials, EMG, EEG, and EOG.
- 2020 – Direct current stimulation boosts hebbian plasticity in vitro
- 2017 – Direct current stimulation alters neuronal input/output function
- 2017 – Direct current stimulation boosts synaptic gain and cooperativity in vitro
- 2016 – Animal models of transcranial direct current stimulation: methods and mechanisms
- 2015 – Reliability of repeated TMS measures in older adults and in patients with subacute and chronic stroke
- 2015 – Modeling sequence and quasi-uniform assumption in computational neurostimulation
- 2015 – Multilevel computational models for predicting the cellular effects of noninvasive brain stimulation
- 2015 – Methods for specific electrode resistance measurement during transcranial direct current stimulation
- 2014 – Clinician accessible tools for GUI computational models of transcranial electrical stimulation: BONSAI and SPHERES
- 2014 – Polarizing cerebellar neurons with transcranial direct current stimulation
- 2013 – Origins of specificity during tDCS: anatomical, activity-selective, and input-bias mechanisms
- 2013 – Cellular effects of acute direct current stimulation: somatic and synaptic terminal effects
- 2013 – The “quasi-uniform” assumption in animal and computational models of non-invasive electrical stimulation
- 2013 – Methods for extra-low voltage transcranial direct current stimulation: current and time dependent impedance decreases
- 2013 – Effects of weak transcranial Alternating Current Stimulation on brain activity–a review of known mechanisms from animal studies
- 2012 – Computational models of transcranial direct current stimulation
- 2012 – Axon terminal polarization induced by weak uniform DC electric fields: a modeling study
- 2012 – Temperature control at DBS electrodes using a heat sink: experimentally validated FEM model of DBS lead architecture
- 2012 – High-resolution modeling assisted design of customized and individualized transcranial direct current stimulation protocols
- 2010 – Low-intensity electrical stimulation affects network dynamics by modulating population rate and spike timing
- 2010 – Electrode montages for tDCS and weak transcranial electrical stimulation: role of “return” electrode’s position and size