Figure 1From: Identification of candidate serum biomarkers for severe septic shock-associated kidney injury via microarrayResults of the leave-one-out cross-validation procedure involving 21 gene probes. The procedure was based on a Support Vector Machines algorithm and was targeted at prediction of 'SSAKI' and 'no SSAKI' classes. Performance calculations provided as the percentage with 95% confidence intervals (CIs). LR, likelihood ratio; NPV, negative predictive value; PPV, positive predictive value; SSAKI, sepsis-shock-associated acute kidney injury.Back to article page