Shengpu Tang

Shengpu Tang

Auto-generated profile, based on publicly available information

Shengpu Tang

Department of Computer Science, College of Arts and Sciences, Emory University, Atlanta, GA, USA

Research Focus & Metrics

Academic Career Length

7Years (2020-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
11 October 2020 in Journal of the American Medical Informatics Association
Verified
Journal Article
Open access
PDF file
STShengpu Tang
Shengpu Tang
PDParmida Davarmanesh
Parmida Davarmanesh
YSYanmeng Song
Yanmeng Song
DKDanai Koutra
Danai Koutra
MSMichael W Sjoding
Michael W Sjoding
JWJenna Wiens
Jenna Wiens
Objective: In applying machine learning (ML) to electronic health record (EHR) data, many decisions must be made before any ML is applied; such preprocessing requires substantial effort and can be labor-intensive. As the role of ML in health care grows, there is an increasing need for systematic and reproducible preprocessing techniques for EHR data. Thus, we developed FIDDLE (Flexible Data-Driven Pipeline), an open-source framework that streamlines the preprocessing of data extracted from the EHR. Materials and Methods: Largely data-driven, FIDDLE systematically transforms structured EHR data into feature vectors, limiting the number of decisions a user must make while incorporating good practices from the literature. To demonstrate its utility and flexibility, we conducted a proof-of-concept experiment in which we applied FIDDLE to 2 publicly available EHR data sets collected from intensive care units: MIMIC-III and the eICU Collaborative Research Database. We trained different ML models to predict 3 clinically important outcomes: in-hospital mortality, acute respiratory failure, and shock. We evaluated models using the area under the receiver operating characteristics curve (AUROC), and compared it to several baselines. Results: Across tasks, FIDDLE extracted 2,528 to 7,403 features from MIMIC-III and eICU, respectively. On all tasks, FIDDLE-based models achieved good discriminative performance, with AUROCs of 0.757–0.886, comparable to the performance of MIMIC-Extract, a preprocessing pipeline designed specifically for MIMIC-III. Furthermore, our results showed that FIDDLE is generalizable across different prediction times, ML algorithms, and data sets, while being relatively robust to different settings of user-defined arguments. Conclusions: FIDDLE, an open-source preprocessing pipeline, facilitates applying ML to structured EHR data. By accelerating and standardizing labor-intensive preprocessing, FIDDLE can help stimulate progress in building clinically useful ML tools for EHR data.

1 November 2020 in JCO Clinical Cancer Informatics
Verified
Journal Article
Open access
STShengpu Tang
Shengpu Tang
GCGrant T. Chappell
Grant T. Chappell
AMAmanda Mazzoli
Amanda Mazzoli
MTMuneesh Tewari
Muneesh Tewari
SCSung Won Choi
Sung Won Choi
JWJenna Wiens
Jenna Wiens
PURPOSE: Acute graft-versus-host disease (aGVHD) remains a significant complication of allogeneic hematopoietic cell transplantation (HCT) and limits its broader application. The ability to predict grade II to IV aGVHD could potentially mitigate morbidity and mortality. To date, researchers have focused on using snapshots of a patient (eg, biomarkers at a single time point) to predict aGVHD onset. We hypothesized that longitudinal data collected and stored in electronic health records (EHRs) could distinguish patients at high risk of developing aGVHD from those at low risk. PATIENTS AND METHODS: The study included a cohort of 324 patients undergoing allogeneic HCT at the University of Michigan C.S. Mott Children’s Hospital during 2014 to 2017. Using EHR data, specifically vital sign measurements collected within the first 10 days of transplantation, we built a predictive model using penalized logistic regression for identifying patients at risk for grade II to IV aGVHD. We compared the proposed model with a baseline model trained only on patient and donor characteristics collected at the time of transplantation and performed an analysis of the importance of different input features. RESULTS: The proposed model outperformed the baseline model, with an area under the receiver operating characteristic curve of 0.659 versus 0.512 ( P = .019). The feature importance analysis showed that the learned model relied most on temperature and systolic blood pressure, and temporal trends (eg, increasing or decreasing) were more important than the average values. CONCLUSION: Leveraging readily available clinical data from EHRs, we developed a machine-learning model for aGVHD prediction in patients undergoing HCT. Continuous monitoring of vital signs, such as temperature, could potentially help clinicians more accurately identify patients at high risk for aGVHD.

1 September 2021 in The American Journal of Surgery
Verified
Journal Article
Closed access
JHJaewon Hur
Jaewon Hur
STShengpu Tang
Shengpu Tang
VGVidhya Gunaseelan
Vidhya Gunaseelan
JVJoceline Vu
Joceline Vu
CBChad M. Brummett
Chad M. Brummett
JWJenna Wiens
Jenna Wiens

28 April 2026 in Nature Medicine
Verified
Journal Article
Closed access
PDF file
WAWouter Anton Christiaan van Amsterdam
Wouter Anton Christiaan van Amsterdam
MOMichael Oberst
Michael Oberst
JFJean Feng
Jean Feng
JWJenna Wiens
Jenna Wiens
STShengpu Tang
Shengpu Tang
AGAnna Goldenberg
Anna Goldenberg

10 October 2025 in JMIR Medical Informatics
Verified
Journal Article
Open access
JLJung Min Lee
Jung Min Lee
STShengpu Tang
Shengpu Tang
MSMichael Sjoding
Michael Sjoding
JWJenna Wiens
Jenna Wiens
Background: Offline reinforcement learning (RL) has been increasingly applied to clinical decision-making problems. However, due to the lack of a standardized pipeline, prior work often relied on strategies that may lead to overfitted policies and inaccurate evaluations. Objective: In this work, we present a practical pipeline—Pipeline for Learning Robust Policies in Reinforcement Learning (PROP-RL)—designed to improve robustness and minimize disruption to clinical workflow. We demonstrate its efficacy in the context of learning treatment policies for administering loop diuretics in hospitalized patients. Methods: Our cohort included adult inpatients admitted to the emergency department at Michigan Medicine between 2015 and 2019 who required supplemental oxygen. We modeled the management of loop diuretics as an offline RL problem using a discrete state space based on features extracted from electronic health records, a binary action space corresponding to the daily use of loop diuretics, and a reward function based on in-hospital mortality. The policy was trained on data from 2015 to 2018 and evaluated on a held-out set of hospitalizations from 2019, in terms of estimated reduction in mortality compared to clinician behavior. Results: The final study cohort included 36,570 hospitalizations. The learned treatment policy was based on 60 states: the policy deferred to clinicians in 36 states, recommended the majority action in 22 states, and diverged significantly from clinician behavior in 2 of the states. Among the cases where the policy meaningfully diverged from the behavior policy, the learned policy was estimated to significantly reduce the mortality rate from 3.8% to 2.2% by 1.6% (95% CI 0.4‐2.7; P=.006). Conclusions: We applied our pipeline to the clinical problem of loop diuretic treatment, highlighting the importance of robust state representation and thoughtful policy selection and evaluation. Our work reveals areas of potential improvement in current clinical care for loop diuretics and serves as a blueprint for using offline RL for sequential treatment selection in clinical settings.

27 August 2026 in SSRN Electronic Journal
Verified
Preprint
Open access
XLXintong Li
Xintong Li
MWMinxiao Wang
Minxiao Wang
RYRunze Yan
Runze Yan
SBStephanie R. Brown
Stephanie R. Brown
MMMark Mai
Mark Mai
STShengpu Tang
Shengpu Tang
Objective: Complex predictive models are increasingly proposed as substitutes for laboratory measurements, yet they are typically assessed with aggregate accuracy metrics that reveal little about downstream clinical impact. We developed an evaluation framework that assesses a prediction model by its utility for guiding decisions under realistic resource constraints, demonstrated through non-invasive lactate estimation in the intensive care unit (ICU).Methods: The framework assumes the model is used to derive a resource allocation policy that selects which ICU patients to re-measure lactate and which to rely on the model's estimate, under a pre-specified budget for invasive blood draws. Policies are evaluated by average clinical cost of errors, area under the receiver operating characteristic curve (AUROC) for hyperlactatemia classification at the clinical threshold, and regression mean squared error (MSE). Beyond random allocation and prediction-ranked or threshold-distance heuristics, we propose Predictive Risk-Informed Sampling (PRIS), which estimates each patient's false-negative and false-positive risk with separate classifiers and ranks by a cost-weighted score, prioritizing draws for cases most vulnerable to harmful misclassification. On MIMIC-III (788 ICU stays) with a pretrained lactate predictor, we compared PRIS against heuristic baselines across budget levels.Results: PRIS significantly reduced average clinical cost versus random allocation at low-to-moderate budgets. All rules improve as measured values replace predictions, but PRIS reaches consequential errors sooner, resolving 22 of 61 misclassifications by 50 draws versus 14 for the prediction-ranked baseline, and achieving the highest classification performance at every budget. Because it ranks by error probability rather than predicted magnitude, PRIS led in clinical cost and classification but was least effective for regression error: the preferred rule depended on the evaluation metric, a distinction that aggregate accuracy would not reveal. At the budget of the routine draw-all-abnormal rule, prediction-ranked identified none of the 34 missed hyperlactatemia cases, whereas PRIS caught 14.Conclusion: Aggregate accuracy does not capture a model's clinical usefulness; evaluating decision utility under a measurement budget is a better alternative. As a triage layer on an existing lactate predictor, PRIS directs lab resources toward cases where inaccurate predictions carry the highest risk of harmful clinical consequences.