Machine Learning–Based Prediction of Delayed Neurologic Sequelae in Carbon Monoxide Poisoning Using Automatically Extracted MR Imaging Features

BACKGROUND AND PURPOSE: Delayed neurologic sequelae are among the most serious complications of carbon monoxide poisoning. However, no reliable tools are available for evaluating their potential risk. We aimed to assess whether machine learning models using imaging features that were automatically extracted from brain MRI can predict the potential delayed neurologic sequelae risk in patients with acute carbon monoxide poisoning. MATERIALS AND METHODS: This single-center, retrospective, observational study analyzed a prospectively collected registry of patients with acute carbon monoxide poisoning who visited our emergency department from April 2011 to December 2015. Overall, 1618 radiomics and 4 lesion-segmentation features from DWI b1000 and ADC images, as well as 62 clinical variables, were extracted from each patient. The entire data set was divided into 5 subsets, with 1 serving as the hold-out test set and the remaining 4 used for training and tuning. Four machine learning models, linear regression, support vector machine, random forest, and extreme gradient boosting, as well as an ensemble model, were trained and evaluated by using 20 different data configurations. The primary evaluation metric was the mean and 95% CI of the area under the receiver operating characteristic curve. Shapley additive explanations were calculated and visualized to enhance model interpretability. RESULTS: Of the 373 patients, delayed neurologic sequelae occurred in 99 (26.5%) patients (mean age 43.0 ± 15.2; 62.0% men). The means [95% CIs] of the area under the receiver operating characteristic curve (AUROC), accuracy, sensitivity, and specificity of the best performing machine learning model for predicting the development of delayed neurologic sequelae were 0.88 [0.86-0.9], 0.82 [0.8-0.83], 0.81 [0.79-0.83], and 0.82 [0.8-0.84], respectively. Among imaging features, the presence, size, and number of acute brain lesions on DWI b1000 and ADC images more accurately predicted delayed neurologic sequelae risk than advanced radiomics features based on shape, texture, and wavelet transformation. CONCLUSIONS: Machine learning models developed using automatically extracted brain MRI features with clinical features can distinguish patients at delayed neurologic sequelae risk. The models enable effective prediction of delayed neurologic sequelae in patients with acute carbon monoxide poisoning, facilitating timely treatment planning for prevention.