Moons, K.G.M.

model patient study regression value disease probability prediction regression coef ficients outcome result predictor infarct size score table treatment stroke history presence roc area analysis population characteristic subject patient history infarct validation diagnosis embolism research ficient approach sensitivity number infarction information level specificity variable meningiti examination visit trial heart disease performance chapter polytomou ratio cancer threshold selection sample difference curve blood teratoma episode selection period method category tissue ischemic ischemic episodes measurement heart prediction rule dichotomou children monitoring prediction models figure practice determinant rotterdam period outcome categories parameter validation sample estimate problem