Classification of Chronic Kidney Disease in Sonography Using the GLCM and Artificial Neural Network

Usefulness of Radiomics and Kidney Volume Based on Non-Enhanced Computed Tomography in Chronic Kidney Disease: Initial Report

INTRODUCTION

Chronic kidney disease (CKD) is classified according to the estimated glomerular filtration rate (eGFR), but kidney volume (KV) can also provide meaningful information. Very few radiomics (RDX) studies on CKD have utilized computed tomography (CT). This study aimed to determine whether non-enhanced computed tomography (NECT)-based RDX can be useful in evaluation of patients with CKD and to compare it with KV.

METHODS

The NECT scans of 64 subjects with impaired kidney function (defined as <60 mL/min/1.73 m2) and 60 controls with normal kidney function were retrospectively analyzed. Kidney segmentations, volume measurements, and RDX features extraction were performed. Machine-learning models using RDX were constructed to classify the kidneys as having structural markers of impaired or normal function.

RESULTS

The median KV in the impaired kidney function group was 114.83 mL vs. 159.43 mL (p < 0.001) in the control group. There was a statistically significant strong positive correlation between KV and eGFR (rs = 0.579, p < 0.001) and a strong negative correlation between KV and serum creatinine level (rs = -0.514, p < 0.001). The KV-based models achieved the best area under the curve (AUC) of 0.746, whereas the RDX-based models achieved the best AUC of 0.878.

CONCLUSIONS

RDX can be useful in identifying patients with impaired kidney function on NECT. RDX-based models outperformed KV-based models. RDX has the potential to identify patients with a higher risk of CKD based on imaging, which, as we believe, can indirectly support clinical decision-making.

A novel clinical-radiomic nomogram for the crescent status in IgA nephropathy

OBJECTIVE

We used machine-learning (ML) models based on ultrasound radiomics to construct a nomogram for noninvasive evaluation of the crescent status in immunoglobulin A (IgA) nephropathy.

METHODS

Patients with IgA nephropathy diagnosed by renal biopsy (n=567) were divided into training (n=398) and test cohorts (n=169). Ultrasound radiomic features were extracted from ultrasound images. After selecting the most significant features using univariate analysis and the least absolute shrinkage and selection operator algorithm, three ML algorithms were assessed for final radiomic model establishment. Next, clinical, ultrasound radiomic, and combined clinical-radiomic models were compared for their ability to detect IgA crescents. The diagnostic performance of the three models was evaluated using receiver operating characteristic curve analysis.

RESULTS

The average area under the curve (AUC) of the three ML radiomic models was 0.762. The logistic regression model performed best, with AUC values in the training and test cohorts of 0.838 and 0.81, respectively. Among the final models, the combined model based on clinical characteristics and the Rad score showed good discrimination, with AUC values in the training and test cohorts of 0.883 and 0.862, respectively. The decision curve analysis verified the clinical practicability of the combined nomogram.

CONCLUSION

ML classifier based on ultrasound radiomics has a potential value for noninvasive diagnosis of IgA nephropathy with or without crescents. The nomogram constructed by combining ultrasound radiomic and clinical features can provide clinicians with more comprehensive and personalized image information, which is of great significance for selecting treatment strategies.

A Catalogue of Machine Learning Algorithms for Healthcare Risk Predictions

Extracting useful knowledge from proper data analysis is a very challenging task for efficient and timely decision-making. To achieve this, there exist a plethora of machine learning (ML) algorithms, while, especially in healthcare, this complexity increases due to the domain's requirements for analytics-based risk predictions. This manuscript proposes a data analysis mechanism experimented in diverse healthcare scenarios, towards constructing a catalogue of the most efficient ML algorithms to be used depending on the healthcare scenario's requirements and datasets, for efficiently predicting the onset of a disease. To this context, seven (7) different ML algorithms (Naïve Bayes, K-Nearest Neighbors, Decision Tree, Logistic Regression, Random Forest, Neural Networks, Stochastic Gradient Descent) have been executed on top of diverse healthcare scenarios (stroke, COVID-19, diabetes, breast cancer, kidney disease, heart failure). Based on a variety of performance metrics (accuracy, recall, precision, F1-score, specificity, confusion matrix), it has been identified that a sub-set of ML algorithms are more efficient for timely predictions under specific healthcare scenarios, and that is why the envisioned ML catalogue prioritizes the ML algorithms to be used, depending on the scenarios' nature and needed metrics. Further evaluation must be performed considering additional scenarios, involving state-of-the-art techniques (e.g., cloud deployment, federated ML) for improving the mechanism's efficiency.

Artificial intelligence in glomerular diseases

In this narrative review, we focus on the application of artificial intelligence in the clinical history of patients with glomerular disease, digital pathology in kidney biopsy, renal ultrasonography imaging, and prediction of chronic kidney disease (CKD). With the development of natural language processing, the clinical history of a patient can be used to identify a computable phenotype. In kidney pathology, digital imaging has adopted innovative deep learning algorithms (DLAs) that can improve the predictive capability of the examined lesions. However, at this time, these applications can only be used in research because there is no recognized validation to replace the conventional diagnostic applications. Kidney ultrasonography, used in the clinical examination of patients, provides information about the progression of kidney damage. Machine learning algorithms (MLAs) with promising results for the early detection of CKD have been proposed, but, still, they are not solid enough to be incorporated into the clinical practice. A few tools for glomerulonephritis, based on MLAs, are available in clinical practice. They can be downloaded on computers and cellular phones but can only be applied to uniracial cohorts of patients. To improve their performance, it is necessary to organize large consortia with multiracial cohorts. Finally, in many studies MLA development has been carried out using retrospective cohorts. The performance of the models might differ in retrospective cohorts compared to real-world data. Therefore, the models should be validated in prospective external large cohorts.

An improved joint non-negative matrix factorization for identifying surgical treatment timing of neonatal necrotizing enterocolitis

Neonatal necrotizing enterocolitis is a severe neonatal intestinal disease. Timely identification of surgical indications is essential for newborns in order to seek the best time for treatment and improve prognosis. This paper attempts to establish an algorithm model based on multimodal clinical data to determine the features of surgical indications and construct an auxiliary diagnosis model. The proposed algorithm adds hypergraph constraints on the two modal data based on Joint Nonnegative Matrix Factorization (JNMF), aiming to mine the higher-order correlations of the two data features. In addition, the adjacency matrix of the two kinds of data is used as a network regularization constraint to prevent overfitting. Orthogonal and L1-norm regulations were introduced to avoid feature redundancy and perform feature selection, respectively, and confirmed 14 clinical features. Finally, we used three classifiers, random forest, support vector machine, and logistic regression, to perform binary classification of patients requiring surgery. The results show that when the features selected by the proposed algorithm model are classified by random forest, the area under the ROC curve is 0.8, which has high prediction accuracy.

Association between urinary metals and leukocyte telomere length involving an artificial neural network prediction: Findings based on NHANES 1999-2002

Objective

Leukocytes telomere length (LTL) was reported to be associated with cellular aging and aging related disease. Urine metal also might accelerate the development of aging related disease. We aimed to analyze the association between LTL and urinary metals.

Methods

In this research, we screened all cycles of National Health and Nutrition Examination Survey (NHANES) dataset, and download the eligible dataset in NHANES 1999-2002 containing demographic, disease history, eight urine metal, and LTL. The analysis in this research had three steps including baseline difference comparison, multiple linear regression (MLR) for hazardous urine metals, and artificial neural network (ANN, based on Tensorflow framework) to make LTL prediction.

Results

The MLR results showed that urinary cadmium (Cd) was negatively correlated with LTL in the USA population [third quantile: -9.36, 95% confidential interval (CI) = (-19.7, -2.32)], and in the elderly urinary molybdenum (Mo) was positively associated with LTL [third quantile: 24.37, 95%CI = (5.42, 63.55)]. An ANN model was constructed, which had 24 neurons, 0.375 exit rate in the first layer, 15 neurons with 0.53 exit rate in the second layer, and 7 neurons with 0.86 exit rate in the third layer. The squared error loss (LOSS) and mean absolute error (MAE) in the ANN model were 0.054 and 0.181, respectively, which showed a low error rate.

Conclusion

In conclusion, in adults especially the elderly, the relationships between urinary Cd and Mo might be worthy of further research. An accurate prediction model based on ANN could be further analyzed.

Diagnosis of Chronic Kidney Disease Using Effective Classification Algorithms and Recursive Feature Elimination Techniques

Chronic kidney disease (CKD) is among the top 20 causes of death worldwide and affects approximately 10% of the world adult population. CKD is a disorder that disrupts normal kidney function. Due to the increasing number of people with CKD, effective prediction measures for the early diagnosis of CKD are required. The novelty of this study lies in developing the diagnosis system to detect chronic kidney diseases. This study assists experts in exploring preventive measures for CKD through early diagnosis using machine learning techniques. This study focused on evaluating a dataset collected from 400 patients containing 24 features. The mean and mode statistical analysis methods were used to replace the missing numerical and the nominal values. To choose the most important features, Recursive Feature Elimination (RFE) was applied. Four classification algorithms applied in this study were support vector machine (SVM), k-nearest neighbors (KNN), decision tree, and random forest. All the classification algorithms achieved promising performance. The random forest algorithm outperformed all other applied algorithms, reaching an accuracy, precision, recall, and F1-score of 100% for all measures. CKD is a serious life-threatening disease, with high rates of morbidity and mortality. Therefore, artificial intelligence techniques are of great importance in the early detection of CKD. These techniques are supportive of experts and doctors in early diagnosis to avoid developing kidney failure.

Towards Accurate Diagnosis of Skin Lesions Using Feedforward Back Propagation Neural Networks

In the automatic detection framework, there have been many attempts to develop models for real-time melanoma detection. To effectively discriminate benign and malign skin lesions, this work investigates sixty different architectures of the Feedforward Back Propagation Network (FFBPN), based on shape asymmetry for an optimal structural design that includes both the hidden neuron number and the input data selection. The reason for the choice of shape asymmetry was based on the 5–10% disagreement between dermatologists regarding the efficacy of asymmetry in the diagnosis of malignant melanoma. Asymmetry is quantified based on lesion shape (contour), moment of inertia of the lesion shape and histograms. The FFBPN has a high architecture flexibility, which indicates it as a favorable tool to avoid the over-parameterization of the ANN and, equally, to discard those redundant input datasets that usually result in poor test performance. The FFBPN was tested on four public image datasets containing melanoma, dysplastic nevus and nevus images. Experimental results on multiple benchmark data sets demonstrate that asymmetry A2 is a meaningful feature for skin lesion classification, and FFBPN with 16 neurons in the hidden layer can model the data without compromising prediction accuracy.