Skip to main content
Fig. 4 | BMC Nephrology

Fig. 4

From: Enhancing individual glomerular filtration rate assessment: can we trust the equation? Development and validation of machine learning models to assess the trustworthiness of estimated GFR compared to measured GFR

Fig. 4

Use of the random forest algorithm for GFR assesment. Under this approach, if the probability of EKFC eGFR falling within 30% of mGFR is less than 80%, mGFR should be measured. The Global P30 metric is then calculated as the ratio of True Positives to the sum of True Negatives and False Negatives. For global P30 with different probability tresholds values, see Table 3. RF: random forest; PPV: positive predictive value; RF: random forest. EKFC: european kidney function consortium; P30EKFC/RF: Probability that EKFC eGFR is within P30 knowing that this probability is superior or equal to the decision treshold, according to the random forest model

Back to article page

BMC Nephrology

ISSN: 1471-2369

Contact us