标签
The paper investigates whether guideline-based categorical encodings of continuous predictors can replace continuous inputs in stroke outcome prediction models without sacrificing accuracy, finding comparable performance in most treatment cohorts.