Tomasz Benedykt Kuliński

Tomasz Benedykt Kuliński

Auto-generated profile, based on publicly available information

Tomasz Benedykt Kuliński

Department of Medical Sciences, Neurosurgery, Uppsala University, Uppsala, Sweden

Research Focus & Metrics

Academic Career Length

1Year (2026-Present)

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

Representative Publications

View All Publications
26 August 2026 in Journal of Cerebral Blood Flow & Metabolism
Verified
Journal Article
Open access
MBMichael S Baker
Michael S Baker
JHJoshua M Heihre
Joshua M Heihre
TKTomasz Kuliński
Tomasz Kuliński
CSClaudia A Smith
Claudia A Smith
BBBerfin Barlas
Berfin Barlas
AHAdel Helmy
Adel Helmy
Cerebral microdialysis (CMD) after traumatic brain injury (TBI) has focused on lactate/pyruvate ratio (LPR), overlooking lactate–pyruvate gradient (LPG, the linear regression slope within a time segment of CMD datapoints) as a proxy for anaerobic activity. We retrospectively applied changepoint detection to the linear lactate–pyruvate relationship (L = m P + b ) over time within patients from two independent CMD-monitored TBI cohorts (Cambridge, n = 510; Uppsala, n = 81), estimating per-segment LPG ( m ) and intercept ( b , capturing metabolic variance). In vitro, we tested whether rotenone-induced mitochondrial dysfunction increases LPG in hiPSC-derived neurons, as expected if LPG tracks anaerobic activity. In the Cambridge cohort, a median of two linear segments were identified per patient (median LPG = 19.83; median b = 0.54 mM). Between segments, LPR tracked LPG with heavy attenuation (β = 0.255). LPG was negatively associated with cerebral microdialysate glucose ( p < 0.001) and 6-month GOSE ( p = 0.039). Uppsala data confirmed pipeline robustness and replicated the LPG–glucose association ( p = 0.022). LPG was higher in rotenone-treated neurons than controls (32.9 vs 14.2, p < 0.001). LPG calculated per time segment provides a more accurate, outcome-associated proxy for anaerobic activity than datapoint-level LPR.

10 September 2026 in Journal of Neurotrauma
Verified
Journal Article
Closed access
MGMarcus Roland Victor Gustafsson
Marcus Roland Victor Gustafsson
TKTomasz Benedykt Kuliński
Tomasz Benedykt Kuliński
DBDavid Bark
David Bark
AWAndre Wyzykowski
Andre Wyzykowski
ERElham Rostami
Elham Rostami
Background: Continuous physiological monitoring in neurocritical care generates large volumes of high-frequency time-series data that remain underutilized in routine traumatic brain injury (TBI) management. Machine learning (ML) approaches applied to these data have emerged as a promising strategy for dynamic prediction of secondary brain injury and individualized clinical decision support. Methods: This narrative review summarizes current ML applications based on continuous physiological monitoring in TBI and related neurocritical care populations. A structured PubMed search supplemented by manual reference screening identified studies published through January 2026. Included studies applied ML methods to continuously acquired physiological signals such as intracranial pressure (ICP), arterial blood pressure (ABP), cerebral perfusion pressure (CPP), and related intensive care unit data. Results: Thirty-four studies were identified, of which 26 met core inclusion criteria focused on continuous physiological monitoring. The most common applications were prediction of intracranial hypertension events, mortality, and functional outcome. Fewer studies addressed arterial hypotension prediction or unsupervised clustering approaches. Recent work demonstrates a transition from static admission-based prognostic models toward dynamic prediction using continuously updated physiological data. Deep learning architectures, including recurrent neural networks and transformers, were increasingly used for high-resolution time-series analysis. However, substantial heterogeneity exists in study design, preprocessing methods, prediction targets, and validation strategies. Most studies relied on retrospective observational data and internal validation, while prospective evaluation and workflow-integrated implementation remain limited. Conclusion: ML approaches using continuous physiological monitoring show significant potential for dynamic prediction and decision support in TBI care. Future progress will depend on multicenter external validation, standardized methodological frameworks, prospective evaluation, and integration into real-time neurocritical care workflows.