Skip to main content
Figure 1 | BMC Medicine

Figure 1

From: A diagnostic algorithm combining clinical and molecular data distinguishes Kawasaki disease from other febrile illnesses

Figure 1

Linear discriminant analysis of KD training cohort patients. (A) The calculated coefficients of linear discriminants (LD1) of each assayed clinical parameter. (B) A modified 2 × 2 contingency table depicting the agreement of classifications with clinical diagnosis. (C) Density plots of the clinical score distribution of both the KD and FC patients. KD, blue density plot; FC, yellow density plot. We stratified all the KD and FC patients, using the clinical scores, into low, intermediate and high groups, where the group boundaries were decided by the diagnosis with 95% accuracy (two dotted vertical lines). Abbreviations: FC, febrile controls; KD, Kawasaki disease.

Back to article page