AKIMLpred: An interpretable machine learning model for predicting acute kidney injury within seven days in critically ill patients based on a prospective cohort study.

BACKGROUND: Early recognition and timely intervention for AKI in critically ill patients were crucial to reduce morbidity and mortality. This study aimed to use biomarkers to construct a optimal machine learning model for early prediction of AKI in critically ill patients within seven days. METHODS: The prospective cohort study enrolled 929 patients altogether who were admitted in ICU including 680 patients in training set (Jiefang Campus) and 249 patients in external testing set (Binjiang Campus). After performing strict inclusion and exclusion criteria, 421 patients were selected in training set for constructing predictive model and 167 patients were selected in external testing for evaluating the predictive performance of resulting model. Urine and blood samples were collected for kidney injury associated biomarkers detection. Baseline clinical information and laboratory data of the study participants were collected. We determined the average prediction efficiency of six machine learning models through 10-fold cross validation. RESULTS: In total, 78 variables were collected when admission in ICU and 43 variables were statistically significant between AKI and non-AKI cohort. Then, 35 variables were selected as independent features for AKI by univariate logistic regression. Spearman correlation analysis was used to remove two highly correlated variables. Three ranking methods were used to explore the influence of 33 variables for further determining the best combination of variables. The gini importance ranking method was found to be applicable for variables filtering. The predictive performance of AKIMLpred which constructed by the XGBoost algorithm was the best among six machine learning models. When the AKIMLpred included the nine features (NGAL, IGFBP7, sCysC, CAF22, KIM-1, NT-proBNP, IL-6, IL-18 and L-FABP) with the highest influence ranking, its model had the best prediction performance, with an AUC of 0.881 and an accuracy of 0.815 in training set, similarly, with an AUC of 0.889 and an accuracy of 0.846 in validation set. Moreover, the performace was slightly outperformed in testing set with an AUC of 0.902 and an accuracy of 0.846. The SHAP algorithm was used to interpret the prediction results of AKIMLpred. The web-calculator of AKIMLpred was shown for predicting AKI with more convenient(https://www.xsmartanalysis.com/model/list/predict/model/html?mid=8065&symbol=11gk693982SU6AE1ms21). AKIMLpred was better than the optimal model built with only routine tests for predicting AKI in critically ill patients within 7 days. CONCLUSION: The model AKIMLpred constructed by the XGBoost algorithm with selecting the nine most influential biomarkers in the gini importance ranking method had the best performance in predicting AKI in critically ill patients within 7 days. This data-driven predictive model will help clinicians to make quick and accurate diagnosis.

Rapid and sensitive detection of L-FABP for prediction and diagnosis of acute kidney injury in critically ill patients by chemiluminescent immunoassay.

BACKGROUND: Acute kidney injury (AKI) was a common clinical complication among critically ill patients in Intensive Care Unit with high morbidity and mortality. Human liver fatty acid-binding protein (L-FABP) as a renal tubular injury biomarker was considered a predictor of AKI; however, high-throughput and sensitive detection methods were still urgently needed. We constructed a sensitive and rapid detection method for detecting L-FABP and for exploring the clinical application of L-FABP as a predictor for AKI. METHODS: We developed an automated detection method of chemiluminescent immunoassay to measure L-FABP and evaluated the analytical performance of the new methodology including analytical selectivity, analytical sensitivity, linear range, the minimum limit of detection (LOD), repeatability, and accuracy. One hundred patients were enrolled in this study to explore the predictive and diagnostic ability for AKI. RESULTS: The chemiluminescent immune-based L-FABP assay had outstanding analytical sensitivity including the detection limit of 0.88 ng/ml, and a wide linear range of 2 ng/ml to 160 ng/ml. It also exhibited excellent repeatability with intra-analysis CVs of 8.73%, 4.72%, and 3.79%, respectively, and the inter-analysis CVs of 13.47%, 7.28%, and 5.94%, respectively. The recovery rate assay exhibited a good accuracy with three L-FABP concentration of 99.76%, 102.27%, and 96.92%, respectively. The reference interval of L-FABP was between 0.88 ng/ml and 5.98 ng/ml. The evaluation of predictive and diagnostic performance showed that higher concentration of L-FABP indicated higher risk of AKI occurrence and disease progression. CONCLUSIONS: The clinical application of rapid and sensitive detection method of L-FABP based on the newly developed chemiluminescent immunoassay could offer benefits for patients. L-FABP was a potentially predictive and diagnostic biomarker for AKI.

Rapid and Quantitative Analysis of Exosomes by a Chemiluminescence Immunoassay Using Superparamagnetic Iron Oxide Particles.

Since the discovery of exosomes, their potential diagnostic value has been the focus of considerable research. However, the lack of a rapid and simple technique for the quantitative analysis of exosomes greatly limits the application of exosomes in clinical research. In this study, we describe a newly developed one-step chemiluminescence immunoassay for the rapid quantitative analysis of exosomes from biofluids. Our new technique, named ExoNANO, adopts a double-antibody sandwich strategy using anti-CD63 antibody-conjugated superparamagnetic iron oxide particles (SIOPs) and acridinium ester (ACE)-labeled anti-CD9 antibodies. SIOPs have narrow size distribution and high magnetic susceptibility, and ACE has excellent chemiluminescent properties such as low background signal and no need for a catalyst. We demonstrated that ExoNANO allows the quantitative analysis of exosomes in the range of 2.92 ×105 to 2.80×108 particles/µL, with a limit of detection of 2.63×105 particles/µL. Using ExoNANO, we quantified exosomes in cell culture medium and clinical biofluids such as serum, saliva, ascitic fluid, and cerebrospinal fluid. We believe that ExoNANO might pave the way for the rapid isolation and quantitative analysis of exosomes for routine clinical applications.