Parthasarathy Thirumala

Associate Professor

University of Pittsburgh Medical Center
Country flagUnited States

Research Interests

ApplyKite TwinLearns as you apply

Make your next move.

Prepare your application

Review drafts before you send them.

Explore related searches

Articles (6)

Objective We developed and validated a detection-guided artifact removal framework for clinical electroencephalography (EEG). The framework applies artifact correction only to the contaminated segments and preserves artifact-free data without modification. Approach The framework uses convolutional neural network (CNN) detectors trained on the Temple University Hospital (TUH) Artifact Corpus (150 recordings, 105 patients). For eye movement artifacts (20 s windows), the framework applies independent component analysis (ICA) and canonical correlation analysis (CCA). For muscle artifacts (5 s windows), the framework applies wavelet thresholding and empirical mode decomposition (EMD). For non-physiological artifacts (1 s windows), the framework applies spherical spline interpolation and artifact subspace reconstruction (ASR). Removal methods were applied only to detector-flagged artifact windows. For windows that are not flagged as artifacts, the output signal is equal to the input signal. In a test set held-out (21 patients, 30 recordings), we compared detection-guided selective removal with global removal, which applies the same method to all windows. We evaluated the preservation of artifact-free windows using correlation, root mean squared error (RMSE), and peak signal-to-noise ratio (PSNR). Main results Selective removal outperformed global removal in all six methods and 18 metric comparisons ( p < 10 −105 ). Selective processing maintained a clean-segment correlation above 0.987 for all methods. Global removal reduced the correlation to values as low as 0.39 (CCA) and 0.47 (ASR). The framework achieved artifact suppression in detector-flagged windows. CCA removed 74.6% of the amplitude of the eye movement artifact. EMD removed 99.8% of high-frequency (30–40 Hz) muscle contamination. ASR reduced the non-physiological artifact amplitude by 37.1% in detector-flagged windows. Preservation of artifact-free windows remained high in all methods, indicating minimal distortion of clean EEG. Significance Detection-guided selective removal addresses a key limitation of pipelines that apply correction globally and that can remove clean electroencephalography (EEG) signals. The framework enables automated artifact removal without manual review and preserves signal fidelity for clinical interpretation. The modular design supports integration into real-time monitoring systems for acute and perioperative care applications.

Year:

2026

Background/Objectives: Intraoperative neurophysiological monitoring (IONM) is used to detect and prevent neurological injury during extramedullary spinal cord tumor (EMSCT) resection, but its diagnostic accuracy lacks systematic validation with recent evidence. This meta-analysis evaluates the performance of somatosensory evoked potentials (SSEPs), transcranial motor evoked potentials (TcMEPs), and multimodal (SSEP + TcMEP) IONM in predicting deficits during EMSCT resections. Methods: Following PRISMA-DTA guidelines, we searched MEDLINE, PubMed, and Ovid (inception to April 2025) for studies on IONM in EMSCT surgeries (PROSPERO: CRD420251047345). Pooled sensitivity, specificity, and reversibility metrics were calculated using bivariate models, with quality assessed via QUADAS-2. Z-test and Bayesian meta-analysis were used for comparisons. Results: Across 20 studies (2672 patients), multimodal IONM showed a log DOR of 4.310 (95% CI: 3.581–5.039) and an AUC of 94.2%, TcMEP monitoring showed a log DOR of 4.367 (95% CI: 3.765–5.127) and an AUC of 92%, while SSEP monitoring showed a log DOR of 3.463 (95% CI: 2.702–4.224) and an AUC of 82%. All modalities demonstrated high specificity (>95%), indicating low false-positive rates. Bayesian analysis revealed >90% probability that TcMEP-based approaches were superior to SSEPs. Reversible TcMEP changes were associated with an 11% (95% CI: 4–24%) postoperative deficit rate, compared to 35% (95% CI: 12–67%) for SSEPs. Conclusions: These findings caution against relying solely on SSEPs and support the use of multimodal IONM strategies, which enhance early detection of impending neurological injury, enable timely surgical interventions, and help prevent permanent neurological damage in EMSCT resections. Although TcMEP and multimodal monitoring showed similar diagnostic accuracy, we continue to recommend multimodal approaches as the current standard of care, pending prospective studies to determine if TcMEP alone can reliably replace multimodal monitoring.

Year:

2025

Objective This study aimed to develop and validate a system of specialized deep lightweight convolutional neural networks (CNN) to accurately detect specific artifact classes and demonstrate their advantage over traditional rule-based methods. Methods Three distinct CNN systems were trained on the Temple University Hospital EEG Artifact Corpus to identify eye movement, muscle-related and non-physiological artifacts, with each system optimized for an ideal temporal window size. The performance of the proposed system was compared with standard rule-based clinical detection methods in a held-out test set. Results The CNN systems significantly outperformed rule-based methods, with F1-score improvements ranging from +11.2% to +44.9%. Importantly, the results revealed distinct optimal temporal window lengths for each artifact type: 20s for eye movements (ROC AUC 0.975%), 5s for muscle activity (Accuracy 93.2%), and 1s for non-physiological artifact(F1-score 77.4%). Conclusion The results show that specialized, artifact-specific CNNs provide a more consistent and accurate solution for automated EEG artifact detection than traditional rule-based approaches Significance This study establishes a new benchmark for automated EEG quality control by validating one of the first open-source, specialized CNN systems for three distinct artifact classes, both high sensitivity and specificity.

Year:

2025

Title Standards for Reporting of Diagnostic Accuracy using Intraoperative Neurophysiological Monitoring (STARD-IONM) Objective Intraoperative neurophysiological monitoring (IONM) plays a critical role in preserving functional integrity during surgery, yet it is challenging to compare studies due to methodological heterogeneity and inconsistent reporting. We developed the STARD-IONM extension to improve the transparency, completeness, and comparability of IONM studies. Methods The STARD-IONM initiative followed a three-phase consensus. Phase 1 convened a IONM expert panel to discuss and define the rationale and scope. Phase 2 involved structured item-level review of existing STARD items in the context of IONM, applied to published studies with iterative feedback. Phase 3 will include broader community engagement via preprints, outreach to professional societies, and public commentary. Results A systematically selected review of IONM studies revealed the underreporting of at key methodological items such as handling missing data (7%), adverse events (11%), and blinding of test and outcomes (22%). A STARD-IONM checklist with recommendations for reporting IONM studies with IONM specific examples were developed. Community feedback emphasized challenges unique to IONM, including the classification of reversible IONM changes, and variability in reference standards. Conclusions The STARD-IONM framework addresses critical gaps in the reporting of diagnostic accuracy studies involving IONM. It represents an application of the STARD criteria, wherein the original checklist has been adapted and supplemented with guidelines for IONM studies. Significance Standardized reporting will facilitate enhanced adherence to methodological standards, increase reproducibility and strengthen the evidence base for the safe and effective use of IONM, which is expected to improve clinical decision-making.

Year:

2025

BACKGROUND AND OBJECTIVES: Intramedullary spinal cord tumor (IMSCT) resection seeks to maximize tumor removal while preserving neurological function, guided by intraoperative neurophysiological monitoring with myogenic motor-evoked potentials (myogenic MEPs) and direct waves (D-wave) recordings. Despite their widespread use, no systematic review has compared their accuracy for predicting immediate, transient, and persistent postoperative motor deficits in IMSCT surgery. This meta-analysis fills this gap by evaluating their diagnostic accuracy. METHODS: Following Preferred Reporting Items for Systematic Reviews and Meta-Analyses for Diagnostic Test Accuracy guidelines, we analyzed 26 studies (1985-2024) involving 1079 patients with IMSCT with myogenic MEP and/or D-wave monitoring. Diagnostic performances were calculated with a bivariate model. RESULTS: For immediate postoperative motor deficits, myogenic MEPs demonstrated higher sensitivity with a log diagnostic odds ratios (log DOR) of 2.91 (95% CI: 2.15-3.67; area under the curve [AUC] 86.1%) compared with D-wave's log DOR 2.00 (95% CI: 0.96-3.04; AUC 77.7%). For transient motor deficits, myogenic MEPs also outperformed D-wave, achieving a log DOR of 1.94 (95% CI: 0.85-3.04; AUC 74.8%) vs D-wave's 0.68 (95% CI: −1.02 to 2.37; AUC 57.4%). For persistent motor deficits, myogenic MEPs maintained higher sensitivity with a robust log DOR of 1.89 (95% CI: 1.22-2.56; AUC 76.7%) and identified 23.1% of cases missed by D-wave in a cohort of 96 dual-monitored patients. However, D-wave showed a superior log DOR of 3.98 (95% CI: 2.55-5.40; AUC 93.4%), underpinned by exceptional specificity. CONCLUSION: Myogenic MEPs and D-wave monitoring play complementary roles in IMSCT resection, requiring appropriate emphasis to optimize resection and neurological outcomes. Consistent with previous studies, myogenic MEPs showed high sensitivity in detecting transient motor deficits, whereas D-wave monitoring remained the most accurate predictor of persistent motor deficits. Our findings further suggest that myogenic MEPs contribute to identifying persistent motor deficits, challenging traditional views on their role. This association, although requiring prospective validation, highlights the need to interpret significant myogenic MEP changes carefully.

Year:

2025

Introduction: Postoperative cognitive decline (POCD) is a common, but often unrecognized condition after surgery. We evaluate postsurgical cognitive changes in a longitudinal population-based study. Methods: The study cohort comprises an age-stratified population-based random sample of individuals aged 65+ years from a small-town region of the United States. We identified as cases those participants who, during annual assessments, self-reported having undergone surgery under general anesthesia during the preceding year. We identified as controls those who reported no surgery for at least 6 consecutive years. The outcome variable changed over time (4 y) in annual cognitive test composite scores encompassing the domains of attention/processing speed, executive function, memory, language, and visuospatial functions, which were modeled using Generalized Estimating Equations. Results: A total of 436 participants had the required cognitive data; 179 participants (cases) had undergone surgery, while 257 participants (controls) had not undergone surgery. After adjusting for age, years of education, gender, race, heart failure, irregular heartbeat, and stroke, participants who underwent surgery had a greater decline in executive function, memory, and attention over a 2-year period. Discussion: Participants who underwent surgery had a significant subsequent decline in cognitive function over the long term.

Year:

2025

Collaborators (10)

Zhaohui Geng

Assistant Professor

Ohio State University

UNITED STATES
View details

Anthony Absalom

Scientific Staff

University Medical Center Groningen

NETHERLANDS
View details

Francesco Sala

Associate Professor

University of Pisa

ITALY
View details

Gea Drost

University Medical Center Groningen

NETHERLANDS
View details

Jay Shils

Professor

Rush University Medical Center

UNITED STATES
View details

Shyam Visweswaran

Chair & Professor

University of Pittsburgh

UNITED STATES
View details

Patrick Bossuyt

Professor of Clinical Epidemiology

Amsterdam University Medical Centers

NETHERLANDS
View details

Marc Nuwer

Distinguished Professor

David Geffen School of Medicine at UCLA

UNITED STATES
View details

Michael G. Fehlings

Robert Campeau Foundation / Dr. C.H. Tator Chair in Brain and Spinal Cord Research at UHN

University of Toronto

CANADA
View details

Kathleen Seidel

University Hospital of Bern

SWITZERLAND
View details