JMIR Publications

ABSTRACT

Background:

Background:

The treatment and care of patients with traumatic brain injury (TBI) constitute an intractable global health problem. Predicting the prognosis and length of hospital stay of patients with TBI may improve therapeutic effects and significantly reduce societal healthcare burden. Applying machine-learning (ML) methods to the field of TBI may be valuable for determining the prognosis and cost-effectiveness of clinical treatment.

Objective:

Objective:

We aimed to use ML methods to build models for predicting the prognosis and length of hospital stay for patients with traumatic craniocerebral injury.

Methods:

Methods:

We collected relevant clinical information from patients treated at the Neurosurgery Centre of the Second Affiliated Hospital of Anhui Medical University between May 2017 and May 2022, of which 80% was used for training the model and 20% for testing via screening and data splitting. We trained and tested the ML models using five cross-validations to avoid overfitting. In the ML models, 11 types of independent variables were used as input variables, and length of stay and GOS score were used as output variables. Once the models were trained, we obtained and compared the errors of each ML model from five rounds of cross-validation to select the best predictive model. The model was then externally tested using clinical data of patients treated at the First Affiliated Hospital of Anhui Medical University from June 2021 to February 2022.

Results:

Results:

The final convolutional neural network-support vector machine (CNN-SVM) model predicted GOS with an accuracy of 93% and 93.69% in the test set and external validation set, respectively. The mean absolute percentage error (MAPE) of the final built convolutional neural network-support vector regression (CNN-SVR) model predicting inpatient time in the test set and external validation set was 10.72% and 10.44%, respectively. The coefficient of determination (R2) was 0.93 and 0.92 in the test set and external validation set, respectively. Compared with back-propagation neural networks (BPNNs), CNN, and SVM models built separately, our hybrid model was identified to be optimal and had high confidence.

Conclusions:

Conclusions:

This study demonstrates the clinical utility of two hybrid models built using ML methods to accurately predict the prognosis and length of hospital stay for patients with TBI. Application of these models may reduce the burden on physicians when assessing TBI and assist clinicians in the medical decision-making process.