Wenhao Zhang 张文豪

Wenhao Zhang 张文豪

Ph.D of Computer Science at University of California, Los Angeles (UCLA).

Publications

Temporal convolutional networks and data rebalancing for clinical length of stay and mortality prediction

Published in Scientific Reports, 2022

It is critical for hospitals to accurately predict patient length of stay (LOS) and mortality in real-time. We evaluate temporal convolutional networks (TCNs) and data rebalancing methods to predict LOS and mortality. This is a retrospective cohort study utilizing the MIMIC-III database. The MIMIC-Extract pipeline processes 24 hour time-series clinical objective data for 23,944 unique patient records. TCN performance is compared to both baseline and state-of-the-art machine learning models including logistic regression, random forest, gated recurrent unit with decay (GRU-D). Models are evaluated for binary classification tasks (LOS>3 days, LOS>7 days, mortality in-hospital, and mortality in-ICU) with and without data rebalancing and analyzed for clinical runtime feasibility. Data is split temporally, and evaluations utilize tenfold cross-validation (stratified splits) followed by simulated prospective hold-out validation. In mortality tasks, TCN outperforms baselines in 6 of 8 metrics (area under receiver operating characteristic, area under precision-recall curve (AUPRC), and F-1 measure for in-hospital mortality; AUPRC, accuracy, and F-1 for in-ICU mortality). In LOS tasks, TCN performs competitively to the GRU-D (best in 6 of 8) and the random forest model (best in 2 of 8). Rebalancing improves predictive power across multiple methods and outcome ratios. The TCN offers strong performance in mortality classification and offers improved computational efficiency on GPU-enabled systems over popular RNN architectures. Dataset rebalancing can improve model predictive power in imbalanced learning. We conclude that temporal convolutional networks should be included in model searches for critical care outcome prediction systems.

Download here

ECG heartbeat classification using deep transfer learning with convolutional neural network and STFT technique

Published in The 4th International Conference on Computing and Data Science, 2022

Best paper award. Electrocardiogram (ECG) is a simple non-invasive measure to identify heart-related issues such as irregular heartbeats known as arrhythmias. While artificial intelligence and machine learning is being utilized in a wide range of healthcare related applications and datasets, many arrhythmia classifiers using deep learning methods have been proposed in recent years. However, sizes of the available datasets from which to build and assess machine learning models is often very small and the lack of well-annotated public ECG datasets is evident. In this paper, we propose a deep transfer learning framework that is aimed to perform classification on a small size training dataset. The proposed method is to fine-tune a general-purpose image classifier ResNet-18 with MIT-BIH arrhythmia dataset in accordance with the AAMI EC57 standard. This paper further investigates many existing deep learning models that have failed to avoid data leakage against AAMI recommendations. We compare how different data split methods impact the model performance. This comparison study implies that future work in arrhythmia classification should follow the AAMI EC57 standard when using any including MIT-BIH arrhythmia dataset.

Download here

Causal Inference in Medicine and in Health Policy: A Summary

Published in HANDBOOK ON COMPUTER LEARNING AND INTELLIGENCE: Volume 2: Deep Learning, Intelligent Control and Evolutionary Computation (page 263-302), 2022

A data science task can be deemed as making sense of the data or testing a hypothesis about it. The conclusions inferred from data can greatly guide us to make informative decisions. Big data has enabled us to carry out countless prediction tasks in conjunction with machine learning, such as identifying high risk patients suffering from a certain disease and taking preventable measures. However, healthcare practitioners are not content with mere predictions – they are also interested in the cause-effect relation between input features and clinical outcomes. Understanding such relations will help doctors treat patients and reduce the risk effectively. Causality is typically identified by randomized controlled trials. Often such trials are not feasible when scientists and researchers turn to observational studies and attempt to draw inferences. However, observational studies may also be affected by selection and/or confounding biases that can result in wrong causal conclusions. In this chapter, we will try to highlight some of the drawbacks that may arise in traditional machine learning and statistical approaches to analyze the observational data, particularly in the healthcare data analytics domain. We will discuss causal inference and ways to discover the cause-effect from observational studies in healthcare domain. Moreover, we will demonstrate the applications of causal inference in tackling some common machine learning issues such as missing data and model transportability. Finally, we will discuss the possibility of integrating reinforcement learning with causality as a way to counter confounding bias.

Download here

Physical Activity Behavior of Patients at a Skilled Nursing Facility: Longitudinal Cohort Study

Published in JMIR mHealth and uHealth, 2022

Background: On-body wearable sensors have been used to predict adverse outcomes such as hospitalizations or fall, thereby enabling clinicians to develop better intervention guidelines and personalized models of care to prevent harmful outcomes. In our previous work, we introduced a generic remote patient monitoring framework (Sensing At-Risk Population) that draws on the classification of human movements using a 3-axial accelerometer and the extraction of indoor localization using Bluetooth low energy beacons, in concert. Using the same framework, this paper addresses the longitudinal analyses of a group of patients in a skilled nursing facility. We try to investigate if the metrics derived from a remote patient monitoring system comprised of physical activity and indoor localization sensors, as well as their association with therapist assessments, provide additional insight into the recovery process of patients receiving rehabilitation.

Download here

A pilot study of a wearable monitoring system as an adjunct to geriatric assessment in older adults with cancer.

Published in Journal of Clinical Oncology, 2020

Background: Advances in health technology provide potential tools that can aid in assessing and monitoring the functional status of the growing older adult population diagnosed with cancer. We piloted a novel wearable monitoring platform, Sensing in At-Risk Populations (SARP), which consists of a smartwatch, software application for health monitoring, and a central data processing and analytics engine.

Download here

Large-scale causal approaches to debiasing post-click conversion rate estimation with multi-task learning

Published in Proceedings of The Web Conference 2020, 2020

Post-click conversion rate (CVR) estimation is a critical task in e-commerce recommender systems. This task is deemed quite chal- lenging under industrial setting with two major issues: 1) selection bias caused by user self-selection, and 2) data sparsity due to the rare click events. A successful conversion typically has the following sequential events: “exposure → click → conversion”. Conventional CVR estimators are trained in the click space, but inference is done in the entire exposure space. They fail to account for the causes of the missing data and treat them as missing at random. Hence, their estimations are highly likely to deviate from the real values by large. In addition, the data sparsity issue can also handicap many indus- trial CVR estimators which usually have large parameter spaces. In this paper, we propose two principled, efficient and highly effective CVR estimators for industrial CVR estimation, namely, Multi-IPW and Multi-DR. The proposed models approach the CVR estimation from a causal perspective and account for the causes of missing not at random. In addition, our methods are based on the multi-task learning framework and mitigate the data sparsity issue. Extensive experiments on industrial-level datasets show that our methods outperform the state-of-the-art CVR models.

Download here

GenSample: A Genetic Algorithm for Oversampling in Imbalanced Datasets

Published in arXiv, 2019

Imbalanced datasets are ubiquitous. Classification performance on imbalanced datasets is generally poor for the minority class as the classifier cannot learn decision boundaries well. However, in sensitive applications like fraud detection, medical diagnosis, and spam identification, it is extremely important to classify the minority instances correctly. In this paper, we present a novel technique based on genetic algorithms, GenSample, for oversampling the minority class in imbalanced datasets. GenSample decides the rate of oversampling a minority example by taking into account the difficulty in learning that example, along with the performance improvement achieved by oversampling it. This technique terminates the oversampling process when the performance of the classifier begins to deteriorate. Consequently, it produces synthetic data only as long as a performance boost is obtained. The algorithm was tested on 9 real-world imbalanced datasets of varying sizes and imbalance ratios. It achieved the highest F-Score on 8 out of 9 datasets, confirming its ability to better handle imbalanced data compared to other existing methodologies.

Download here

WOTBoost: Weighted Oversampling Technique in Boosting for imbalanced learning

Published in 5th Special Session on Intelligent Data Mining in IEEE BigData 2019, 2019

Machine learning classifiers often stumble over imbalanced datasets where classes are not equally represented. This inherent bias towards the majority class may result in low accuracy in labeling minority class. Imbalanced learning is prevalent in many real-world applications, such as medical research, network intrusion detection, and fraud detection in credit card transactions, etc. A good number of research works have been reported to tackle this challenging problem. For example, Synthetic Minority Over-sampling TEchnique (SMOTE) and ADAptive SYNthetic sampling approach (ADASYN) use oversampling techniques to balance the skewed datasets. In this paper, we propose a novel method that combines a Weighted Oversampling Technique and ensemble Boosting method (WOTBoost) to improve the classification accuracy of minority data without sacrificing the accuracy of the majority class. WOTBoost adjusts its oversampling strategy at each round of boosting to synthesize more targeted minority data samples. The adjustment is enforced using a weighted distribution. We compare WOTBoost with other four classification models (i.e., decision tree, SMOTE + decision tree, ADASYN + decision tree, SMOTEBoost) extensively on 18 public accessible imbalanced datasets. WOTBoost achieves the best G mean on 6 datasets and highest AUC score on 7 datasets.

Download here

A Combination of Indoor Localization and Wearable Sensor–Based Physical Activity Recognition to Assess Older Patients Undergoing Subacute Rehabilitation: Baseline Study Results

Published in JMIR mHealth and uHealth, 2019

Background: Health care, in recent years, has made great leaps in integrating wireless technology into traditional models of care. The availability of ubiquitous devices such as wearable sensors has enabled researchers to collect voluminous datasets and harness them in a wide range of health care topics. One of the goals of using on-body wearable sensors has been to study and analyze human activity and functional patterns, thereby predicting harmful outcomes such as falls. It can also be used to track precise individual movements to form personalized behavioral patterns, to standardize the concept of frailty, well-being/independence, etc. Most wearable devices such as activity trackers and smartwatches are equipped with low-cost embedded sensors that can provide users with health statistics. In addition to wearable devices, Bluetooth low-energy sensors known as BLE beacons have gained traction among researchers in ambient intelligence domain. The low cost and durability of newer versions have made BLE beacons feasible gadgets to yield indoor localization data, an adjunct feature in human activity recognition. In the studies by Moatamed et al and the patent application by Ramezani et al, we introduced a generic framework (Sensing At-Risk Population) that draws on the classification of human movements using a 3-axial accelerometer and extracting indoor localization using BLE beacons, in concert.

Download here