From: Early prediction of chronic kidney disease based on ensemble of deep learning models and optimizers
Paper | Detection/prediction | Dataset/number of samples | Dataset description | Algorithm | Highest accuracy |
---|---|---|---|---|---|
Qin-2019 [11] | Detection | Not available | Demographic features, Laboratory results, and ultrasound images | Logistic regression, random forest, support vector machines | 99.75% |
Jongbo-2020 [1] | Detection | CKD dataset-1/400 [15] | Demographic features: age, sex, race, and ethnicity. Medical features: blood pressure, blood sugar | Ensemble: KNN, NB, DT | 100% |
Ma-2020 [13] | Detection | Not available | Ultrasound images | SVM, DT, RF, KNN, HMANN | HMANN 98% |
Gudeti-2020 [7] | Detection | CKD dataset-1/400 [15] | Demographic features: age, sex, race, and ethnicity. Medical features: blood pressure, blood sugar | SVM, LR and KNN | SVM 99.2% |
Chittora-2021 [12] | Detection | CKD dataset-1/400 [15] | Demographic features: age, sex, race, and ethnicity. Medical features: blood pressure, blood sugar | C5.0, CHAID, ANN, LSVM, LR, RT and KNN | LSVM 98.86% |
Senan-2021 [8] | Detection | CKD dataset-1/400 [15] | Demographic features: age, sex, race, and ethnicity. Medical features: blood pressure, blood sugar | SVM, RF, KNN, DT | RF 100% |
Alsuhibany-2021[14] | Detection | CKD dataset-1/400 [15] | Demographic features: age, sex, race, and ethnicity. Medical features: blood pressure, blood sugar | Ensemble (DBN, KELM, CNN-GRU) | %96.9 |
Krishnamurthy-2021 [9] | Prediction | CKD dataset-2/90,000 [10] | Comorbidities, medications, age, gender, | CNN, BLSTM, LightGBM, LR, RF, DT | CNN |
89% (6 months) | |||||
88% (12 months) | |||||
Singh-2022 [17] | Detection | CKD dataset-1/400 [15] | Demographic features: age, sex, race, and ethnicity. Medical features: blood pressure, blood sugar | SVM, KNN, LR, RF, Naive Bayes and DNN | DNN 100% |
Sawhney-2023 [18] | Detection | Not available | age, blood sugar, red blood cell counts and medical data | SVM, DT, MLP, RF, LR | MLP 100% |