Figure 3
From: Identification of the robust predictor for sepsis based on clustering analysis
Comparison between three and four clusters established by K-means clustering. Colored rectangles indicate cases with prominently high number of common subjects between two subgroups. (A) Results are described number and column-based ratio for subjects including healthy control and sepsis patients in common between two subgroups. (B) Results are described as number of sepsis patients/healthy cases and ratio [sepsis patients/(sepsis + healthy subjects)].
