About

Scott A. Bruce is an Associate Professor of Statistics at Texas A&M University. His research combines methodological and collaborative statistics, with a focus on nonstationary and multivariate time series, adaptive frequency band analysis, Bayesian spectral modeling, and scalable statistical learning for complex data. His work is motivated by transdisciplinary problems in sleep and circadian science, cardiology, psychiatry, neuroscience, biomechanics, and digital health. He develops practical, computationally efficient methods for extracting interpretable structure from high-dimensional signals, wearable measurements, and longitudinal observations. His methodological focus includes frequency-domain methods for stationary, nonstationary, multivariate, and functional time series, Bayesian approaches for adaptive spectral estimation and covariate-dependent spectra, data-driven frequency band construction for interpretability, and statistical learning tools for high-dimensional and heterogeneous data. His research areas encompass biomedical signal analysis, sleep and circadian science, cardiovascular biomarkers, psychiatry, neuroscience, biomechanics, and digital health. His recent projects involve software development for frequency band analysis, adaptive frequency band learning in functional time series, classification of categorical time series, and covariate-dependent spectral analysis using Bayesian tree ensembles. His work aims to develop interpretable, adaptive, and computationally feasible methods for analyzing temporally structured data arising in modern observational and experimental studies.

Research topics

• Medicine
• Internal medicine
• Composite material
• Cardiology
• Pathology
• Environmental chemistry
• Biology
• Bioinformatics
• Materials science
• Nanotechnology
• Demography
• Environmental health
• Immunology
• Virology
• Chemistry

Selected publications

• Frequency Band Analysis of Multiple Stationary Time Series

  Statistics in Medicine · 2026-02-01

  articleOpen accessSenior author

  The frequency-domain properties of biomedical signals offer valuable insights into health and functioning of underlying physiological systems. The power spectrum, which characterizes these properties, is often summarized by partitioning frequencies into standard bands and averaging power within bands. These summary measures are regularly used for analysis in practice, but are not guaranteed to optimally retain differences in power spectra across signals from different participants. We propose a data-adaptive method for identifying frequency band summary measures that preserve spectral variability within a population of interest. The method can also identify subpopulations with distinct power spectra and summary measures that best characterize local dynamics. Numerical selection criteria are developed to select a reasonable number of bands and subpopulations that best characterize overall dynamics. A genetic algorithm is designed to simultaneously identify subpopulations and their corresponding summary measures. The method is used to analyze stride interval series from patients with different neurological disorders, revealing distinct subpopulations and the need for subpopulation-dependent summary measures.

  PublisherDOI

• sbruce23/RehabHFproteomics: Initial release

  Zenodo (CERN European Organization for Nuclear Research) · 2026-01-15

  otherOpen access1st authorCorresponding

  No description provided.

  PublisherDOI

• sbruce23/RehabHFproteomics: Initial release

  Zenodo (CERN European Organization for Nuclear Research) · 2026-01-15

  otherOpen access1st authorCorresponding

  No description provided.

  PublisherDOI

• Circulating Biomarkers as Predictors of Improvement in Physical Function in Hospitalized Older Adults With Geriatric Syndromes: Findings From the REHAB-HF Trial

  Circulation Heart Failure · 2026-01-07

  article

  BACKGROUND: Biomarkers in heart failure (HF) provide mechanistic and prognostic insights, but their role in predicting treatment response is less understood. We evaluated whether multiple baseline biomarker profiles from the REHAB-HF trial (Rehabilitation Therapy in Older Acute Heart Failure Patients) could stratify functional improvement following a 12-week physical rehabilitation intervention (RI). METHODS: Participants ≥60 years hospitalized with heart failure were randomized to a 12-week outpatient RI or attention control. Functional outcomes included changes in the short physical performance battery and 6-minute walk distance. Blood collected at baseline and 12 weeks was analyzed for cardiac (cTnI and cTnT, NT-proBNP [N-terminal pro-brain natriuretic peptide]), renal (creatinine), and inflammatory (CRP [C-reactive protein]) biomarkers. Associations between baseline biomarker levels and 12-week functional gains by treatment group were evaluated using adjusted linear regression models and machine learning-based decision trees. RESULTS: =0.040 and 0.032). In the decision tree, analyses combining all biomarkers, CRP emerged as the primary biomarker for both outcomes. Among participants with CRP ≥9.9 mg/L, RI was associated with a +2.4 point (95% CI, 1.8-3.1) greater increase in the short physical performance battery and a +79 m (95% CI, 50-109) greater increase in 6-minute walk distance compared with attention control. In contrast, for those with CRP <9.9 mg/L, the differential benefit of the RI was limited (+0.8 in short physical performance battery [95% CI, 0.1-1.6]; +30 m in 6-minute walk distance [95% CI, -1.0 to 61]). The biomarker levels (except for creatinine) decreased by 12 weeks posthospitalization, but with no differences based on treatment assignment. CONCLUSIONS: Higher inflammation, measured by CRP, may identify older adults recently hospitalized for heart failure with the greatest functional benefit from a physical RI. Biomarker profiling may predict the benefits of this treatment. REGISTRATION: URL: https://www.clinicaltrials.gov; Unique identifier: NCT02196038.

  PublisherDOI

• Idealized Framework for Assisting Pharmacovigilance Reporting in an Ambulatory Primary Care and Chronic Disease Management Clinic

  Pharmacoepidemiology · 2025-11-21

  articleOpen access

  Pharmacovigilance approaches have conventionally focused on the use of epidemiological data to detect emergent adverse drug reactions (ADRs). Recent advances in the use and availability of real-world data have expanded opportunities to detect ADR signals in medical records. We provide a limited review of pharmacovigilance practices and tools we have specifically considered implementing into our comprehensive medication management clinic and associated research programs. Use of pharmacogenomic variants has proven useful only on a limited scale as such data are reliant on low-dimensional approaches matching variants to drugs, often with small effect sizes. As such, most ADRs go unrecognized, undocumented, and unactionable. We posit that an idealized pharmacovigilance framework that relies on artificial-intelligence-assisted reporting with adjudication by pharmacovigilance experts and new models of ambulatory pharmaceutical practice would establish the following attributes: (1) all metadata relating to medication use would be available in the medical record in computable and interoperable data models, (2) digital surveillance tools would detect most ADR events with attributed pharmacological contributions, (3) all events would be characterized using standard adjudication rubrics, and (4) all events would iteratively inform an ADR knowledgebase and improve models to advance detection and prediction of ADR during the course of patient care with a focus on having the necessary tools for clinicians to prevent ADRs. This review provides a limited and focused framework for more systematic documentation of ADRs and tactics to mitigate the idiopathic nature of most ADRs.

  PublisherOA PDFDOI

• TREATMENT WITH TORSEMIDE REDUCES NT-PROBNP COMPARED TO FUROSEMIDE IN PATIENTS DISCHARGED AFTER AN ACUTE HEART FAILURE HOSPITALIZATION

  Journal of the American College of Cardiology · 2025-03-29

  articleOpen access
  PublisherDOI

• Lung Transplant Referral for Central Airway Stenosis and Bronchiectasis Secondary to Lye Inhalation Injury

  American Journal of Respiratory and Critical Care Medicine · 2025-05-01

  article

  Abstract Introduction: Lye (sodium hydroxide or potassium hydroxide) is a strong alkali that can cause caustic burns upon ingestion. We describe a unique patient with lye-associated tracheal stenosis who underwent LT evaluation at our center. Description: A 41-year-old woman with a history of bronchiectasis was referred to our center for lung transplant evaluation. At the time of presentation, she was tracheostomized (uncuffed Shiley 6.0), on room air at baseline, and had a gastrojejunal (GJ) tube. She had a history of intentional lye ingestion with Pop Rocks at age 16. Subsequently, she had endured a supraglottic tear that was deemed irreparable at the time. She was tracheostomized and underwent several tracheal dilations for tracheal stenosis. She also developed esophageal stenosis and needed a GJ tube. The patient had also experienced episodes of recurrent infection and aspiration events leading to further airway damage. On initial consultation, the patient was independent at baseline and walked 453 meters in 6 minutes with lowest oxygen saturation of 90%. Bronchoscopy demonstrated tracheitis with plaques (image), and imaging demonstrated bronchiectasis with patent tracheobronchial tree; cultures grew Pseudomonas aeruginosa. Given no recent exacerbations and an acceptable quality of life, she was deemed too early for transplant consideration and referred to the Pulmonary Clinic. Discussion: Our case illustrates the relationship between severe airway damage from lye ingestion and the potential for lung transplantation. The pathophysiology of lye injury is secondary to saponification of the chemical resulting in tissue necrosis, cellular damage, and deep, extensive burns leading to fibrosis and scarring resulting in tracheal and esophageal injury. Mediastinal fibrosis may pose a higher surgical risk. While lung transplantation remains a potential option, careful consideration of the extent of airway damage, the patient's overall nutritional status, infection control, and pulmonary function is paramount to optimizing outcomes.

  PublisherDOI

• Frequency band analysis of nonstationary multivariate time series

  Biometrics · 2025-07-03

  articleOpen accessSenior author

  Information from frequency bands in biomedical time series provides useful summaries of the observed signal. Many existing methods consider summaries of the time series obtained over a few well-known, pre-defined frequency bands of interest. However, there is a dearth of data-driven methods for identifying frequency bands that optimally summarize frequency-domain information in the time series. A new method to identify partition points in the frequency space of a multivariate locally stationary time series is proposed. These partition points signify changes across frequencies in the time-varying behavior of the signal and provide frequency band summary measures that best preserve nonstationary dynamics of the observed series. An $L_2$-norm based discrepancy measure that finds differences in the time-varying spectral density matrix is constructed, and its asymptotic properties are derived. New nonparametric bootstrap tests are also provided to identify significant frequency partition points and to identify components and cross-components of the spectral matrix exhibiting changes over frequencies. Finite-sample performance of the proposed method is illustrated via simulations. The proposed method is used to develop optimal frequency band summary measures for characterizing time-varying behavior in resting-state electroencephalography time series, as well as identifying components and cross-components associated with each frequency partition point.

  PublisherOA PDFDOI

• Persistent Leukocytosis During Lung Transplant Evaluation: Unmasking Myelodysplastic Syndrome!

  2024-04-30

  article1st authorCorresponding
  PublisherDOI

• ANOPOW for replicated nonstationary time series in experiments

  The Annals of Applied Statistics · 2024-01-31 · 3 citations

  articleOpen accessSenior author

  We propose a novel analysis of power (ANOPOW) model for analyzing replicated nonstationary time series commonly encountered in experimental studies. Based on a locally stationary ANOPOW Cramér spectral representation, the proposed model can be used to compare the second-order time-varying frequency patterns among different groups of time series and to estimate group effects as functions of both time and frequency. Formulated in a Bayesian framework, independent two-dimensional second-order random walk (RW2D) priors are assumed on each of the time-varying functional effects for flexible and adaptive smoothing. A piecewise stationary approximation of the nonstationary time series is used to obtain localized estimates of time-varying spectra. Posterior distributions of the time-varying functional group effects are then obtained via integrated nested Laplace approximations (INLA) at a low computational cost. The large-sample distribution of local periodograms can be appropriately utilized to improve estimation accuracy since INLA allows modeling of data with various types of distributions. The usefulness of the proposed model is illustrated through two real data applications: analyses of seismic signals and pupil diameter time series in children with attention deficit hyperactivity disorder. Simulation studies, Supplementary Materials (Li, Yue and Bruce, 2023a), and R code (Li, Yue and Bruce, 2023b) for this article are also available.

  PublisherOA PDFDOI

Recent grants

• Adaptive Frequency Band Estimation and Analysis

  NIH · $1.2M · 2020–2026

Frequent coauthors

• Anne White

  9 shared

• J. G. Ratcliffe

  9 shared

• Andrew Swift

  9 shared

• Zeda Li

  9 shared

• S. Dobson

  NHS Lothian

  9 shared

• Brani Vidaković

  8 shared

• Dixon Vimalajeewa

  8 shared

• Christopher R. deFilippi

  6 shared

Education

• Ph.D. Statistics, Statistical Science

  Temple University

  2018

• M.S. Statistics, Industrial and Systems Engineering

  Georgia Institute of Technology

  2010

• B.S. Industrial and Systems Engineering, Industrial and Systems Engineering

  Georgia Institute of Technology

  2009