Improving accuracy of estimating glomerular filtration rate using artificial neural network: model development and validation.
Ontology highlight
ABSTRACT:
SUBMITTER: Li N
PROVIDER: S-EPMC7063770 | biostudies-literature | 2020 Mar
REPOSITORIES: biostudies-literature
Publications
Improving accuracy of estimating glomerular filtration rate using artificial neural network: model development and validation.
Li Ningshan N Huang Hui H Qian Han-Zhu HZ Liu Peijia P Lu Hui H Liu Xun X
Journal of translational medicine 20200310 1
<h4>Background</h4>The performance of previously published glomerular filtration rate (GFR) estimation equations degrades when directly used in Chinese population. We incorporated more independent variables and using complicated non-linear modeling technology (artificial neural network, ANN) to develop a more accurate GFR estimation model for Chinese population.<h4>Methods</h4>The enrolled participants came from the Third Affiliated Hospital of Sun Yat-sen University, China from Jan 2012 to Jun ...[more]
Similar Datasets
Project description:BackgroundAccurate evaluation of glomerular filtration rates (GFRs) is of critical importance in clinical practice. A previous study showed that models based on artificial neural networks (ANNs) could achieve a better performance than traditional equations. However, large-sample cross-sectional surveys have not resolved questions about ANN performance.MethodsA total of 1,180 patients that had chronic kidney disease (CKD) were enrolled in the development data set, the internal validation data set and the external validation data set. Additional 222 patients that were admitted to two independent institutions were externally validated. Several ANNs were constructed and finally a Back Propagation network optimized by a genetic algorithm (GABP network) was chosen as a superior model, which included six input variables; i.e., serum creatinine, serum urea nitrogen, age, height, weight and gender, and estimated GFR as the one output variable. Performance was then compared with the Cockcroft-Gault equation, the MDRD equations and the CKD-EPI equation.ResultsIn the external validation data set, Bland-Altman analysis demonstrated that the precision of the six-variable GABP network was the highest among all of the estimation models; i.e., 46.7 ml/min/1.73 m(2) vs. a range from 71.3 to 101.7 ml/min/1.73 m(2), allowing improvement in accuracy (15% accuracy, 49.0%; 30% accuracy, 75.1%; 50% accuracy, 90.5% [P<0.001 for all]) and CKD stage classification (misclassification rate of CKD stage, 32.4% vs. a range from 47.3% to 53.3% [P<0.001 for all]). Furthermore, in the additional external validation data set, precision and accuracy were improved by the six-variable GABP network.ConclusionsA new ANN model (the six-variable GABP network) for CKD patients was developed that could provide a simple, more accurate and reliable means for the estimation of GFR and stage of CKD than traditional equations. Further validations are needed to assess the ability of the ANN model in diverse populations.
| S-EPMC3596400 | biostudies-literature
Project description:BackgroundGlomerular filtration rate (GFR) is essential for renal function evaluation and classification of chronic kidney disease (CKD), while the reference method in children are cumbersome. In the Chinese children, there was no data about GFR measured through plasma or renal clearance of the exogenous markers, and therefore no validated GFR-estimating tools in this population.MethodsWe measured GFR with double-sample plasma clearance of 99mTc-DTPA (mGFR) in 87 hospitalized children with renal injury. Using mGFR as the golden standard, we evaluate the efficiency of four different GFR estimation equations (the original and update Schwartz equation, the Filler equation, the CKiD equation) by statistical parameters of correlation, precision and accuracy.ResultsIn our population, mGFR was 97.0± 31.9 mL/min/1.73m2. The updated Schwartz equation, the Filler equation and the CKiD equation, produced eGFR with strong correlation with mGFR, strong explanation capacity of variance in mGFR, small bias, satisfactory performance in Bland-Altman analysis, high intra-class correlation coefficients, high ratio of eGFR within mGFR±10% and eGFR within mGFR±30%, good agreement in CKD staging between eGFR and mGFR. The original Schwartz equation produced eGFR with large bias, poor precision and accuracy.ConclusionThe validated equations to estimate GFR in our patients are the updated Schwartz equation, which is simple for bedside use, the Filler equation and the CKiD equation, which provide more accurate eGFR. The original Schwartz equation should not be applied to estimate GFR in Chinese children with kidney injuries.
| S-EPMC5500331 | biostudies-literature
Project description:BackgroundAccurate assessment of kidney function is clinically important, but estimates of glomerular filtration rate (GFR) by regression are imprecise.MethodsWe hypothesized that ensemble learning could improve precision. A total of 1419 participants were enrolled, with 1002 in the development dataset and 417 in the external validation dataset. GFR was independently estimated from age, sex and serum creatinine using an artificial neural network (ANN), support vector machine (SVM), regression, and ensemble learning. GFR was measured by 99mTc-DTPA renal dynamic imaging calibrated with dual plasma sample 99mTc-DTPA GFR.ResultsMean measured GFRs were 70.0 ml/min/1.73 m2 in the developmental and 53.4 ml/min/1.73 m2 in the external validation cohorts. In the external validation cohort, precision was better in the ensemble model of the ANN, SVM and regression equation (IQR = 13.5 ml/min/1.73 m2) than in the new regression model (IQR = 14.0 ml/min/1.73 m2, P < 0.001). The precision of ensemble learning was the best of the three models, but the models had similar bias and accuracy. The median difference ranged from 2.3 to 3.7 ml/min/1.73 m2, 30% accuracy ranged from 73.1 to 76.0%, and P was > 0.05 for all comparisons of the new regression equation and the other new models.ConclusionsAn ensemble learning model including three variables, the average ANN, SVM, and regression equation values, was more precise than the new regression model. A more complex ensemble learning strategy may further improve GFR estimates.
| S-EPMC5679185 | biostudies-literature
Project description:Glomerular filtration is a pivotal process of renal physiology, and its alterations are a central pathological event in acute kidney injury and chronic kidney disease. Creatinine clearance (ClCr), a standard method for glomerular filtration rate (GFR) measurement, requires a long and tedious procedure of timed (usually 24 h) urine collection. We have developed a neural network (NN)-based calculator of rat ClCr from plasma creatinine (pCr) and body weight. For this purpose, matched pCr, weight, and ClCr trios from our historical records on male Wistar rats were used. When evaluated on the training (1165 trios), validation (389), and test sets (660), the model committed an average prediction error of 0.196, 0.178, and 0.203 mL/min and had a correlation coefficient of 0.863, 0.902, and 0.856, respectively. More importantly, for all datasets, the NN seemed especially effective at comparing ClCr among groups within individual experiments, providing results that were often more congruent than those measured experimentally. ACLARA, a friendly interface for this calculator, has been made publicly available to ease and expedite experimental procedures and to enhance animal welfare in alignment with the 3Rs principles by avoiding unnecessary stressing metabolic caging for individual urine collection.
| S-EPMC8945373 | biostudies-literature
Project description:BackgroundNew creatinine-based estimated glomerular filtration rate (eGFR) equations, including the 2021 Chronic Kidney Disease Epidemiology Collaboration (2021 CKD-EPI) and European Kidney Function Consortium (EKFC) equations, have been introduced recently. We assessed the performance of the 2021 CKD-EPI and EKFC equations in the Korean population.MethodsWe analyzed 1,654 Korean patients aged ≥18 years who underwent chromium-51-ethylenediamine tetraacetic acid GFR measurements (mGFR). Bias (eGFR-mGFR), root mean square error (RMSE), and proportion of eGFR within 30% of mGFR (P30) of the 2009 CKD-EPI, 2021 CKD-EPI, and EFKC equations were compared. The concordance rate between eGFR and mGFR categories was evaluated. Both eGFR and mGFR categories were classified into six groups: ≥90, 89-60, 59-45, 44-30, 29-15, and <15 mL/min/1.73 m2.ResultsThe median bias (mL/min/1.73 m2) was 1.8 for the 2009 CKD-EPI equation, 4.8 for the 2021 CKD-EPI equation, and -0.3 for the EKFC equation. The P30 and RMSE were 78.2% and 17.0 for the 2009 CKD-EPI equation, 75.6% and 17.4 for the 2021 CKD-EPI equation, and 80.0% and 16.7 for the EKFC equation, respectively. The overall GFR category concordance rate between eGFR and mGFR was 63.4% for the 2009 CKD-EPI equation, 60.5% for the 2021 CKD-EPI equation, and 61.0% for the EKFC equation.ConclusionsAmong the three eGFR equations, the EKFC equation had the smallest bias and highest P30 in Koreans. The 2009 CKD-EPI equation had a lower bias than the 2021 CKD-EPI equation.
| S-EPMC9791020 | biostudies-literature
Project description:BackgroundWe aimed to assess the performance of the five creatinine-based equations commonly used for estimates of the glomerular filtration rate (eGFR), namely, the creatinine-based Chronic Kidney Disease Epidemiology Collaboration (CKD-EPIcr), Asian CKD-EPI, revised Lund-Malmö (revised LM), full age spectrum (FAS), and Korean FAS equations, in the Korean population.MethodsA total of 1,312 patients, aged 20 yr and above who underwent ⁵¹Cr-EDTA GFR measurements (mGFR), were enrolled. The bias (eGFR-mGFR) and precision (root mean square error [RMSE]) were calculated. The accuracy (P30) of four eGFR equations was compared to that of the CKD-EPIcr equation. P30 was defined as the percentage of patients whose eGFR was within±30% of the mGFR.ResultsThe mean bias (mL·min⁻¹·1.73 m⁻²) of the five eGFR equation was as follows: CKD-EPIcr, -0.6; Asian CKD-EPI, 2.7; revised LM, -6.5; FAS, -2.5; and Korean FAS, -0.2. The bias of the Asian CKD-EPI, revised LM, and FAS equations showed a significant difference from zero (P<0.001). The RMSE values were as follows: CKD-EPIcr, 15.6; Asian CKD-EPI, 15.6; revised LM, 17.9; FAS, 16.3; and Korean FAS, 15.8. There were no significant differences in the P30 except for the Asian CKD-EPI equation: CKD-EPIcr, 76.6%; Asian CKD-EPI, 74.7%; revised LM, 75.8%; FAS, 76.0%; and Korean FAS, 75.8%.ConclusionsThe CKD-EPIcr and Korean FAS equations showed equivalent analytical and clinical performances in the Korean adult population.
| S-EPMC5500735 | biostudies-literature
Project description:IntroductionGlomerular filtration rate (GFR) is typically estimated with equations that use biomarkers such as serum creatinine and/or cystatin-C. The impact of these different biomarkers on GFR estimates in glomerular disease patients is unclear. In this study, we compared the different GFR estimating equations in the Cure Glomerulonephropathy (CureGN) cohort of children and adults with glomerular disease.MethodsAll available cystatin-C measurements from CureGN study participants were matched to same-day serum creatinine measurements to estimate GFR. To explore the strength of agreement between eGFR values obtained from the "Under 25" (U25) and Chronic Kidney Disease Epidemiology Collaboration (CKD-Epi) equations, we used intraclass correlation coefficients. Multivariable linear mixed effects models were used to determine which factors were independently associated with differences in eGFR values.ResultsA total of 928 cystatin-C measurements were matched to same-day serum creatinine measurements from N = 332 CureGN study participants (58% male, 69% White/Caucasian, 20% Black/African American). Among 628 measurements collected while study participants were under 25 years old, there was moderate agreement (0.731) in serum creatinine versus cystatin-C U25 equations. Models showed that higher eGFR values were associated with larger differences between the two equations (p < 0.001). Among 253 measurements collected while study participants were at least 18 years old, there was excellent agreement (0.891-0.978) among CKD-Epi equations using serum creatinine alone, cystatin-C alone, or the combination of both. Younger age was associated with larger differences between CKD-Epi equations (p = 0.06 to p = 0.016).ConclusionExcellent agreement between CKD-Epi equations indicates continued use of serum creatinine alone for GFR estimation could be appropriate for adults. In contrast, only moderate agreement between U25 equations indicates a need for more frequent measurement of cystatin-C among children and young adults, especially as eGFR increases.
| S-EPMC11250572 | biostudies-literature
Project description:ImportanceThere is increasing concern that continued use of a glomerular filtration rate (GFR) estimating equation adjusted for a single racial group could exacerbate chronic kidney disease-related disparities and inequalities.ObjectiveTo assess the performance of GFR estimating equations across varied patient populations.Data sourcesPubMed, Embase, Web of Science, ClinicalTrials.gov, and Scopus databases were systematically searched from January 2012 to February 2023.Study selectionInclusion criteria were studies that compared measured GFR with estimated GFR in adults using established reference standards and methods. A total of 6663 studies were initially identified for screening and review.Data extraction and synthesisFollowing Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines, 2 authors independently extracted data on studies that examined the bias and accuracy of GFR estimating equations. For each outcome, a random-effects model was used to calculate pooled estimates. Data analysis was conducted from March to December 2023.Main outcomes and measuresThe primary outcomes were bias and accuracy of estimated GFRs in Black vs non-Black patients, as well as in individuals with chronic conditions. Bias was defined as the median difference between the measured GFR and the estimated GFR. Accuracy was assessed with P30 (the proportion of persons in a data set whose estimated GFR values were within 30% of measured GFR values) and measures of heterogeneity.ResultsA total of 12 studies with a combined 44 721 patients were included. Significant heterogeneity was found in the bias of various GFR estimation equations. Race-corrected equations and creatinine-based equations tended to overestimate GFR in Black populations and showed mixed results in non-Black populations. For creatinine-based equations, the mean bias in subgroup analysis was 2.1 mL/min/1.73 m2 (95% CI, -0.2 mL/min/1.73 m2 to 4.4 mL/min/1.73 m2) in Black persons and 1.3 mL/min/1.73 m2 (95% CI, 0.0 mL/min/1.73 m2 to 2.5 mL/min/1.73 m2) in non-Black persons. Equations using only cystatin C had small biases. Regarding accuracy, heterogeneity was high in both groups. The overall P30 was 84.5% in Black persons and 87.8% in non-Black persons. Creatinine-based equations were more accurate in non-Black persons than in Black persons. For creatinine-cystatin C equations, the P30 was higher in non-Black persons. There was no significant P30 difference in cystatin C-only equations between the 2 groups. In patients with chronic conditions, P30 values were generally less than 85%, and the biases varied widely.Conclusions and relevanceThis systematic review and meta-analysis of GFR estimating equations suggests that there is bias in race-based GFR estimating equations, which exacerbates kidney disease disparities. Development of a GFR equation independent of race is a crucial starting point, but not the sole solution. Addressing the disproportionate burden of kidney failure on Black individuals in the US requires an enduring, multifaceted approach that should include improving diagnostics, tackling social determinants of health, confronting systemic racism, and using effective disease prevention and management strategies.
| S-EPMC10915689 | biostudies-literature
Project description:Estimates of glomerular filtration rate (GFR) that are based on serum creatinine are routinely used; however, they are imprecise, potentially leading to the overdiagnosis of chronic kidney disease. Cystatin C is an alternative filtration marker for estimating GFR.Using cross-sectional analyses, we developed estimating equations based on cystatin C alone and in combination with creatinine in diverse populations totaling 5352 participants from 13 studies. These equations were then validated in 1119 participants from 5 different studies in which GFR had been measured. Cystatin and creatinine assays were traceable to primary reference materials.Mean measured GFRs were 68 and 70 ml per minute per 1.73 m(2) of body-surface area in the development and validation data sets, respectively. In the validation data set, the creatinine-cystatin C equation performed better than equations that used creatinine or cystatin C alone. Bias was similar among the three equations, with a median difference between measured and estimated GFR of 3.9 ml per minute per 1.73 m(2) with the combined equation, as compared with 3.7 and 3.4 ml per minute per 1.73 m(2) with the creatinine equation and the cystatin C equation (P=0.07 and P=0.05), respectively. Precision was improved with the combined equation (interquartile range of the difference, 13.4 vs. 15.4 and 16.4 ml per minute per 1.73 m(2), respectively [P=0.001 and P<0.001]), and the results were more accurate (percentage of estimates that were >30% of measured GFR, 8.5 vs. 12.8 and 14.1, respectively [P<0.001 for both comparisons]). In participants whose estimated GFR based on creatinine was 45 to 74 ml per minute per 1.73 m(2), the combined equation improved the classification of measured GFR as either less than 60 ml per minute per 1.73 m(2) or greater than or equal to 60 ml per minute per 1.73 m(2) (net reclassification index, 19.4% [P<0.001]) and correctly reclassified 16.9% of those with an estimated GFR of 45 to 59 ml per minute per 1.73 m(2) as having a GFR of 60 ml or higher per minute per 1.73 m(2).The combined creatinine-cystatin C equation performed better than equations based on either of these markers alone and may be useful as a confirmatory test for chronic kidney disease. (Funded by the National Institute of Diabetes and Digestive and Kidney Diseases.).
| S-EPMC4398023 | biostudies-other
Project description:The glomerular filtration rate (GFR) estimating equation incorporating both cystatin C and creatinine perform better than those using creatinine or cystatin C alone in patients with reduced GFR. Whether this equation performs well in kidney transplant recipients cross-sectionally, and more importantly, over time has not been addressed.We analyzed four GFR estimating equations in participants of the Angiotensin II Blockade for Chronic Allograft Nephropathy Trial (NCT 00067990): Chronic Kidney Disease Epidemiology Collaboration equations based on serum cystatin C and creatinine (eGFR (CKD-EPI-Creat+CysC)), cystatin C alone (eGFR (CKD-EPI-CysC)), creatinine alone (eGFR (CKD-EPI-Creat)) and the Modification of Diet in Renal Disease study equation (eGFR (MDRD)). Iothalamate GFR served as a standard (mGFR).mGFR, serum creatinine, and cystatin C shortly after transplant were 56.1 ± 17.0 ml/min/1.73 m(2), 1.2 ± 0.4 mg/dl, and 1.2 ± 0.3 mg/l respectively. eGFR (CKD-EPI-Creat+CysC) was most precise (R(2) = 0.50) but slightly more biased than eGFR (MDRD); 9.0 ± 12.7 versus 6.4 ± 15.8 ml/min/1.73 m(2), respectively. This improved precision was most evident in recipients with mGFR >60 ml/min/1.73 m(2). For relative accuracy, eGFR (MDRD) and eGFR (CKD-EPI-Creat+CysC) had the highest percentage of estimates falling within 30% of mGFR; 75.8 and 68.9%, respectively. Longitudinally, equations incorporating cystatin C most closely paralleled the change in mGFR.eGFR (CKD-EPI-Creat+CysC) is more precise and reflects GFR change over time reasonably well. eGFR (MDRD) had superior performance in recipients with mGFR between 30 and 60 ml/min/1.73 m(2).
| S-EPMC4026030 | biostudies-literature