Predict, diagnose, and treat chronic kidney disease with machine learning: a systematic literature review

Sanmarchi, Francesco; Fanconi, Claudio; Golinelli, Davide; Gori, Davide; Hernandez‐Boussard, Tina; Capodici, Angelo

doi:10.1007/s40620-023-01573-4

Cited by 41 publications

(21 citation statements)

References 98 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…At present, some known risk factors affecting the progression of CKD mainly include age, sex ratio, smoking history, BMI, diabetes, hypertension, and family history of kidney disease [ 15 ]. Besides, there are many other risk factors affecting the development of kidney disease, such as proteinuria, infection, and autoimmune diseases [ 16 ]. Hence, many scholars try to predict the progression of CKD based on specific variables by developing statistical models.…”

Section: Discussionmentioning

confidence: 99%

Risk prediction of the progression of chronic kidney disease stage 1 based on peripheral blood samples: construction and internal validation of a nomogram

Wu,

Zhou,

et al. 2023

Renal Failure

View full text Add to dashboard Cite

Patients with chronic kidney disease (CKD) have high morbidity and mortality, and the disease progression has a significant impact on their survival and living standards. This research aims to analyze risk factors for CKD stage 1 and provide a reference for clinical decision making. The clinical data and peripheral blood samples of 300 patients with CKD stage 1 were collected retrospectively. Patients were randomly assigned into a training set ( n = 210) and a validation set ( n = 90). Patients’ baseline characteristic levels were subjected to statistical tests for difference. Univariate and multivariate Cox regression analyses were utilized to identify risk factors influencing disease progression. Subsequently, a prediction model for disease progression was developed using a nomogram, and the model’s accuracy was assessed using the C-index and calibration curve. The results revealed that hypertension, diabetes, and urinary albumin were essential factors in the progression of CKD stage 1. The nomogram was constructed and then the C-index was calculated. The calibration curve was utilized to assess the risk model. The C-index of the training set was 0.75, and the C-index of the validation set was 0.73, suggesting a good predictive ability of the model. The risk model accurately predicted the progression of CKD stage 1, which is of great significance to developing personalized treatment for patients in clinical practice.

show abstract

Section: Discussionmentioning

confidence: 99%

Risk prediction of the progression of chronic kidney disease stage 1 based on peripheral blood samples: construction and internal validation of a nomogram

Wu,

Zhou,

et al. 2023

Renal Failure

View full text Add to dashboard Cite

show abstract

“…Machine learning, which has shown success in predicting diseases [9], holds promise within nephrology [10] including enhancing CKD screening and detection [3,[11][12][13]. One promising application of machine learning in the context of CKD is the potential for at-home detection or screening.…”

Section: Introductionmentioning

confidence: 99%

Machine learning for classifying chronic kidney disease and predicting creatinine levels using at-home measurements

Metherall,

Berryman,

Brennan

2024

Preprint

View full text Add to dashboard Cite

Background: Chronic kidney disease (CKD) is a global health concern with early detection playing a pivotal role in effective management. Machine learning models demonstrate promise in CKD detection, yet the impact on detection and classification using different sets of clinical features remains under-explored. Methods: In this study, we focus on CKD classification and creatinine prediction using three sets of features; at-home, monitoring, and laboratory. We employ artificial neural networks (ANNs) and random forests (RFs) on a dataset of 400 patients with 25 input features, which we divide into three feature sets. Using 10-fold cross-validation, we calculate metrics such as accuracy, true positive rate (TPR), true negative rate (TNR), and mean squared error. Results: Our results reveal RF achieves superior accuracy (92.5\%) in at-home CKD classification over ANNs (82.9\%). ANNs achieve a higher TPR (92.0\%) but a lower TNR (67.9\%) compared with RFs (90.0\% and 95.8\%, respectively). For monitoring and laboratory features, both methods achieve accuracies exceeding 98\%. The R2 score for creatinine regression is approximately 0.3 higher with laboratory features than at-home features. Feature importance analysis identifies key clinical variables hemoglobin and blood urea, and key comorbidities hypertension and diabetes mellitus, in agreement with previous studies. Conclusions: Machine learning models, particularly RFs, exhibit promise in CKD diagnosis and highlight significant features in CKD detection. Moreover, such models may assist in screening a general population using at-home features---potentially increasing early detection of CKD, thus improving patient care and offering hope for a more effective approach to managing this prevalent health condition.

show abstract

“…Literature analysis shows that the same kidney diseases of different types are often diagnosed in diametrically opposite ways [14], [15]. For example, primary non-obstructive pyelonephritis (PL) usually initially manifests with general symptoms, followed by local symptoms after 2-3 days.…”

Section: Introductionmentioning

confidence: 99%

Development of mathematical methods for diagnosing kidney diseases using fuzzy set tools

Myrzakerimova,

Kolesnikova

2024

IJEECS

View full text Add to dashboard Cite

An approach based on fuzzy set theory is presented in the scientific article to enhance the efficiency of diagnosing kidney diseases by decreasing the time required for making medical decisions. The suggested approach employs fuzzy models and algorithms that consider the uncertainty and variability of clinical data to optimize the assessment of the functional state of the kidneys, taking into account various risk factors and individual characteristics of patients. The paper suggests a technique to develop a system of fuzzy decision rules. This technique combines E. Shortliff’s iterative rules with functions from the studied classes of kidney diseases. Mathematical modeling and experimental studies have indicated relatively high effectiveness in classifying different forms of kidney diseases. The results can be used to formulate intelligent decision support systems in clinical practice and improve diagnostic and monitoring processes. Moreover, the findings may aid in shaping more targeted and effective health policies at the national and regional levels, enhancing access to healthcare, and promoting the population’s overall health.

show abstract

Predict, diagnose, and treat chronic kidney disease with machine learning: a systematic literature review

Cited by 41 publications

References 98 publications

Risk prediction of the progression of chronic kidney disease stage 1 based on peripheral blood samples: construction and internal validation of a nomogram

Risk prediction of the progression of chronic kidney disease stage 1 based on peripheral blood samples: construction and internal validation of a nomogram

Machine learning for classifying chronic kidney disease and predicting creatinine levels using at-home measurements

Development of mathematical methods for diagnosing kidney diseases using fuzzy set tools

Contact Info

Product

Resources

About