Discovery of Generalizable TBI Phenotypes Using Multivariate Time-Series Clustering.

Area of Science:

• Neuroscience
• Computational Biology
• Medical Informatics

Background:

• Traumatic Brain Injury (TBI) exhibits significant heterogeneity, complicating patient stratification and treatment.
• Existing TBI phenotyping studies often lack generalizability across diverse clinical settings and populations.
• Identifying consistent TBI phenotypes is crucial for advancing personalized medicine and improving patient outcomes.

Purpose of the Study:

• To develop and validate a robust method for identifying generalizable Traumatic Brain Injury (TBI) phenotypes.
• To uncover dynamic patterns and distinct clinical profiles within heterogeneous TBI patient cohorts.
• To assess the stability and applicability of identified phenotypes across different datasets.

Main Methods:

• Employed a self-supervised learning-based approach for clustering multivariate time-series data with missing values (SLAC-Time).
• Analyzed two large-scale datasets: the research-centric TRACK-TBI and the real-world MIMIC-IV.
• Utilized multivariate time-series clustering to identify dynamic TBI characteristics.

Main Results:

• The SLAC-Time method demonstrated consistent optimal hyperparameters and cluster numbers across heterogeneous datasets, confirming its stability.
• Identified three generalizable TBI phenotypes (α, β, and γ) with distinct clinical presentations during emergency department visits and temporal profiles during ICU stays.
• Phenotype α characterized mild TBI, phenotype β severe TBI with diverse manifestations, and phenotype γ moderate TBI.
• Age was a significant factor in TBI outcomes, with older individuals experiencing higher mortality, though core phenotype characteristics remained consistent across age groups.

Conclusions:

• The study successfully identified three stable and generalizable TBI phenotypes using a novel computational approach.
• These phenotypes offer a more unified framework for understanding TBI heterogeneity across diverse patient populations and clinical settings.
• The findings pave the way for more targeted therapeutic strategies and improved management of Traumatic Brain Injury.