From Admission to Discharge: Predicting National Institutes of Health Stroke Scale Progression in Stroke Patients Using Biomarkers and Explainable Machine Learning

Abstract

As a result of social progress and improved living conditions, which have contributed to a prolonged life expectancy, the prevalence of strokes has increased and has become a significant phenomenon. Despite the available stroke treatment options, patients frequently suffer from significant disability after a stroke. Initial stroke severity is a significant predictor of functional dependence and mortality following an acute stroke. The current study aims to collect and analyze data from the hyperacute and acute phases of stroke, as well as from the medical history of the patients, in order to develop an explainable machine learning model for predicting stroke-related neurological deficits at discharge, as measured by the National Institutes of Health Stroke Scale (NIHSS). More specifically, we approached the data as a binary task problem: improvement of NIHSS progression vs. worsening of NIHSS progression at discharge, using baseline data within the first 72 h. For feature selection, a genetic algorithm was applied. Using various classifiers, we found that the best scores were achieved from the Random Forest (RF) classifier at the 15 most informative biomarkers and parameters for the binary task of the prediction of NIHSS score progression. RF achieved 91.13% accuracy, 91.13% recall, 90.89% precision, 91.00% f1-score, 8.87% FNrate and 4.59% FPrate. Those biomarkers are: age, gender, NIHSS upon admission, intubation, history of hypertension and smoking, the initial diagnosis of hypertension, diabetes, dyslipidemia and atrial fibrillation, high-density lipoprotein (HDL) levels, stroke localization, systolic blood pressure levels, as well as erythrocyte sedimentation rate (ESR) levels upon admission and the onset of respiratory infection. The SHapley Additive exPlanations (SHAP) model interpreted the impact of the selected features on the model output. Our findings suggest that the aforementioned variables may play a significant role in determining stroke patients’ NIHSS progression from the time of admission until their discharge.

Automated Extraction of Stroke Severity from Unstructured Electronic Health Records using Natural Language Processing

“…Other studies have predicted stroke severity by approaching the data as a binary task problem [32,33], and thus are not directly comparable to ours. In one study [33], a 3D-Convolutional…”

“…Other studies have predicted stroke severity by approaching the data as a binary task problem [32,33], and thus are not directly comparable to ours. In one study [33], a 3D-Convolutional Neural Network was trained to predict low (NIHSS < 5) vs. high (NIHSS ≥ 5) based on preprocessed diffusion-weighted imaging (DWI) images.…”

“…However, limiting to a binary classification limits the utility of the model, especially in light of a NIHSS of 6 being put in the same category as an NIHSS of 42. In a different study [32], the authors developed a random forests model that achieved a recall, precision and F1-score of 0.91 to predict improvement vs. worsening of NIHSS progression as a binary outcome from hospital admission until discharge, using baseline data within the first 72 hours of admission. By predicting NIHSS as a continuous variable, and enabling the prediction of NIHSS even when it is not documented or is missing from the EHR, our model overcomes the limitations of these prior works.…”

“…First, we split the MGH data randomly into a training set (70%) and hold-out test set (30%), as in previous studies [30][31][32]. With the training data, we developed a linear regression model using the least absolute shrinkage and selection operator (LASSO) that utilized the text-based features from the notes to predict the patients' NIHSS scores.…”