Machine learning algorithms for predicting outcomes of traumatic brain injury: A systematic review and meta-analysis

Abstract

Background: Traumatic brain injury (TBI) is a leading cause of death and disability worldwide. The use of machine learning (ML) has emerged as a key advancement in TBI management. This study aimed to identify ML models with demonstrated effectiveness in predicting TBI outcomes. Methods: We conducted a systematic review in accordance with the Preferred Reporting Items for Systematic Review and Meta-Analysis statement. In total, 15 articles were identified using the search strategy. Patient demographics, clinical status, ML outcome variables, and predictive characteristics were extracted. A small meta-analysis of mortality prediction was performed, and a meta-analysis of diagnostic accuracy was conducted for ML algorithms used across multiple studies. Results: ML algorithms including support vector machine (SVM), artificial neural networks (ANN), random forest, and Naïve Bayes were compared to logistic regression (LR). Thirteen studies found significant improvement in prognostic capability using ML versus LR. The accuracy of the above algorithms was consistently over 80% when predicting mortality and unfavorable outcome measured by Glasgow Outcome Scale. Receiver operating characteristic curves analyzing the sensitivity of ANN, SVM, decision tree, and LR demonstrated consistent findings across studies. Lower admission Glasgow Coma Scale (GCS), older age, elevated serum acid, and abnormal glucose were associated with increased adverse outcomes and had the most significant impact on ML algorithms. Conclusion: ML algorithms were stronger than traditional regression models in predicting adverse outcomes. Admission GCS, age, and serum metabolites all have strong predictive power when used with ML and should be considered important components of TBI risk stratification.

Fluid-Based Protein Biomarkers in Traumatic Brain Injury: The View from the Bedside

“…One approach, with increasing interest in the literature, is a range of statistical approaches under the umbrella of Big Data Analytics [184] or Artificial Intelligence (AI) and machine learning [7,185,186]. There are a wide variety of mathematical models that can be utilized to try and identify occult relationships between variables using brute force statistical approaches from large aggregates of data.…”

“…These are tall orders, but all the necessary components are available: several potential mechanistic fluid-based protein biomarkers have already been identified [3]. Moreover, all the required technologies are readily available including analytical platforms enabling sensitive and multiplexed assays [4][5][6], as well as the necessary algorithms for data processing and machine learning (ML) [7]. However, to help to incorporate blood-based 2 of 22 protein biomarker data into a patient management tool, two critical dimensions of TBI need to be considered, the spectrum of disease severity and complexity and the time factor.…”

A Prospective Longitudinal Cohort Study of Serum Stanniocalcin-1 as a Potential Prognostic Biomarker of Severe Traumatic Brain Injury

“… 2 Severe TBI (sTBI) accounts for 8% of all TBI worldwide, with approximately 5.48 million people suffering from sTBI annually. 3 sTBI mortality is from 20% to 30% and its prognosis is mainly related to trauma severity. 3 Secondary brain injury following sTBI involves hemorrhagic and ischemic cerebral injury, and the specific pathophysiological mechanisms include inflammatory reaction, mitochondrial dysfunction, cortical spreading depression, oxidative stress, microvascular thrombosis, neuronal necrosis and apoptosis, brain edema and blood–brain barrier disruption.…”

“… 3 sTBI mortality is from 20% to 30% and its prognosis is mainly related to trauma severity. 3 Secondary brain injury following sTBI involves hemorrhagic and ischemic cerebral injury, and the specific pathophysiological mechanisms include inflammatory reaction, mitochondrial dysfunction, cortical spreading depression, oxidative stress, microvascular thrombosis, neuronal necrosis and apoptosis, brain edema and blood–brain barrier disruption. 4 Conventionally, the Glasgow coma scale (GCS) is selected as a clinical severity indicator, which can be used to discriminate the risk of poor clinical outcome of sTBI.…”

Machine Learning Algorithm Predicts Mortality Risk in Intensive Care Unit for Patients with Traumatic Brain Injury

“…Courville E et al reported a systematic view and meta-analysis (2013–2020) demonstrating that much of this literature discusses in-hospital mortality and poor prognosis, but lacks a more specific focus on the ICU population to understand the predictive power of AIs in TBI patients [ 17 ]. In the last three years, there have been several reports on the prognosis and mortality risk of brain injury using ML techniques.…”