Prediction of Intracranial Hypertension and Brain Tissue Hypoxia Utilizing High-Resolution Data from the BOOST-II Clinical Trial

Abstract

The current approach to intracranial hypertension and brain tissue hypoxia is reactive, based on fixed thresholds. We used statistical machine learning on high-frequency intracranial pressure (ICP) and partial brain tissue oxygen tension (PbtO 2 ) data obtained from the BOOST-II trial with the goal of constructing robust quantitative models to predict ICP/PbtO 2 crises. We derived the following machine learning models: logistic regression (LR), elastic net, and random forest. We split the data set into 70–30% for training and testing and utilized a discrete-time survival analysis framework and 5-fold hyperparameter optimization strategy for all models. We compared model performances on discrimination between events and non-events of increased ICP or low PbtO 2 with the area under the receiver operating characteristic (AUROC) curve. We further analyzed clinical utility through a decision curve analysis (DCA). When considering discrimination, the number of features, and interpretability, we identified the RF model that combined the most recent ICP reading, episode number, and longitudinal trends over the preceding 30 min as the best performing for predicting ICP crisis events within the next 30 min (AUC 0.78). For PbtO 2 , the LR model utilizing the most recent reading, episode number, and longitudinal trends over the preceding 30 min was the best performing (AUC, 0.84). The DCA showed clinical usefulness for wide risk of thresholds for both ICP and PbtO 2 predictions. Acceptable alerting thresholds could range from 20% to 80% depending on a patient-specific assessment of the benefit-risk ratio of a given intervention in response to the alert.

Multi-Mechanistic Approaches to the Treatment of Traumatic Brain Injury: A Review

“…Several recent groups have advocated for the use of machine learning and “big data” analytic approaches to synthesize this information and classify patients into distinct physiologic states, allowing for an individualized yet systematic approach to treating the evolving physiologic derangements caused by TBI [ 13 , 204 , 205 ]. For example, while the BOOST-II trial was not statistically powered to guide outcomes-oriented treatment, a machine learning analysis of BOOST-II data used a combination of logistic regression, elastic net, and random forest machine learning methods to derive clinically applicable predictive models for ICP and brain oxygenation that could be used for early intervention and treatment of intracranial hypertension and hypoxia [ 206 ]. Moving forward, future work linking large high-fidelity data sets of MMM-derived physiologic data with long-term clinical outcomes could be used to further drive advances in TBI treatment [ 188 , 207 ].…”

Artificial Intelligence and Machine Learning Applications in Critically Ill Brain Injured Patients

“…For instance, using data from the Brain Tissue Oxygen Monitoring and Management in Severe Traumatic Brian Injury (BOOST) II trial, ML predicted both ICP spikes and brain tissue hypoxia with a 30-minute lead time. 62 However, the optimal ML algorithms for forecasting ICP and PbtO 2 differed, highlighting that often an integrated approach may be advantageous when dealing with multiple physiological parameters.…”

Multi-mechanistic Approaches to the Treatment of Traumatic Brain Injury: A Review

“…Several recent groups have advocated for the use of machine learning and "big data" analytic approaches to synthesize this information and classify patients into distinct physiologic states, allowing for an individualized yet systematic approach to treating the evolving physiologic derangements caused by TBI [11,168,169]. For example, while the BOOST-II trial was not statistically powered to guide outcomes-oriented treatment, a machine learning analysis of BOOST-II data used a combination of logistic regression, elastic net and random forest machine learning methods to derive clinically applicable predictive models for ICP and brain oxygenation that could be used for early intervention and treatment of intracanial hypertension and hypoxia [170]. Moving forward, future work linking large high-fidelity data sets of MMM-derived physiologic data with long term clinical outcomes could be used to further drive advances in TBI treatment [151,171].…”