DTI-based radiomics signature for the detection of early diabetic kidney damage

Diffusion Tensor and Kurtosis MRI-Based Radiomics Analysis of Kidney Injury in Type 2 Diabetes

BACKGROUND

Diffusion tensor imaging (DTI) and diffusion kurtosis imaging (DKI) can provide quantitative parameters that show promise for evaluation of diabetic kidney disease (DKD). The combination of radiomics with DTI and DKI may hold potential clinical value in detecting DKD.

PURPOSE

To investigate radiomics models of DKI and DTI for predicting DKD in type 2 diabetes mellitus (T2DM) and evaluate their performance in automated renal parenchyma segmentation.

STUDY TYPE

Prospective.

POPULATION

One hundred and sixty-three T2DM patients (87 DKD; 63 females; 27-80 years), randomly divided into training cohort (N = 114) and validation cohort (N = 49).

FIELD STRENGTH/SEQUENCE

1.5-T, diffusion spectrum imaging (DSI) with 9 different b-values.

ASSESSMENT

The images of DSI were processed to generate DKI and DTI parameter maps, including fractional anisotropy (FA), mean diffusivity (MD), axial diffusivity (AD), and radial diffusivity (RD). The Swin UNETR model was trained with 5-fold cross-validation using 100 samples for renal parenchyma segmentation. Subsequently, radiomics features were automatically extracted from each parameter map. The performance of the radiomics models on the validation cohort was evaluated by utilizing the receiver operating characteristic (ROC) curve.

STATISTICAL TESTS

Mann-Whitney U test, Chi-squared test, Pearson correlation coefficient, least absolute shrinkage and selection operator (LASSO), dice similarity coefficient (DSC), decision curve analysis (DCA), area under the curve (AUC), and DeLong's test. The threshold for statistical significance was set at P < 0.05.

RESULTS

The DKI_MD achieved the best segmentation performance (DSC, 0.925 ± 0.011). A combined radiomics model (DTI_FA, DTI_MD, DKI_FA, DKI_MD, and DKI_RD) showed the best performance (AUC, 0.918; 95% confidence interval [CI]: 0.820-0.991). When the threshold probability was greater than 20%, the combined model provided the greatest net benefit. Among the single parameter maps, the DTI_FA exhibited superior diagnostic performance (AUC, 887; 95% CI: 0.779-0.972).

DATA CONCLUSION

The radiomics signature constructed based on DKI and DTI may be used as an accurate and non-invasive tool to identify T2DM and DKD.

LEVEL OF EVIDENCE

2 TECHNICAL EFFICACY: Stage 2.

Ultrasound-Based Radiomics for the Classification of Henoch-Schönlein Purpura Nephritis in Children

Henoch-Schönlein purpura nephritis (HSPN) is one of the most common kidney diseases in children. The current diagnosis and classification of HSPN depend on pathological biopsy, which is seriously limited by its invasive and high-risk nature. The aim of the study was to explore the potential of radiomics model for evaluating the histopathological classification of HSPN based on the ultrasound (US) images. A total of 440 patients with Henoch-Schönlein purpura nephritis proved by biopsy were analyzed retrospectively. They were grouped according to two histopathological categories: those without glomerular crescent formation (ISKDC grades I-II) and those with glomerular crescent formation (ISKDC grades III-V). The patients were randomly assigned to either a training cohort (n = 308) or a validation cohort (n = 132) with a ratio of 7:3. The sonologist manually drew the regions of interest (ROI) on the ultrasound images of the right kidney including the cortex and medulla. Then, the ultrasound radiomics features were extracted using the Pyradiomics package. The dimensions of radiomics features were reduced by Spearman correlation coefficients and least absolute shrinkage and selection operator (LASSO) method. Finally, three radiomics models using k-nearest neighbor (KNN), logistic regression (LR), and support vector machine (SVM) were established, respectively. The predictive performance of such classifiers was assessed with receiver operating characteristic (ROC) curve. 105 radiomics features were extracted from derived US images of each patient and 14 features were ultimately selected for the machine learning analysis. Three machine learning models including k-nearest neighbor (KNN), logistic regression (LR), and support vector machine (SVM) were established for HSPN classification. Of the three classifiers, the SVM classifier performed the best in the validation cohort [area under the curve (AUC) =0.870 (95% CI, 0.795-0.944), sensitivity = 0.706, specificity = 0.950]. The US-based radiomics had good predictive value for HSPN classification, which can be served as a noninvasive tool to evaluate the severity of renal pathology and crescentic formation in children with HSPN.

Current progress in artificial intelligence-assisted medical image analysis for chronic kidney disease: A literature review

Chronic kidney disease (CKD) causes irreversible damage to kidney structure and function. Arising from various etiologies, risk factors for CKD include hypertension and diabetes. With a progressively increasing global prevalence, CKD is an important public health problem worldwide. Medical imaging has become an important diagnostic tool for CKD through the non-invasive identification of macroscopic renal structural abnormalities. Artificial intelligence (AI)-assisted medical imaging techniques aid clinicians in the analysis of characteristics that cannot be easily discriminated by the naked eye, providing valuable information for the identification and management of CKD. Recent studies have demonstrated the effectiveness of AI-assisted medical image analysis as a clinical support tool using radiomics- and deep learning-based AI algorithms for improving the early detection, pathological assessment, and prognostic evaluation of various forms of CKD, including autosomal dominant polycystic kidney disease. Herein, we provide an overview of the potential roles of AI-assisted medical image analysis for the diagnosis and management of CKD.

Diagnostic accuracy of ultrasound-based multimodal radiomics modeling for fibrosis detection in chronic kidney disease

OBJECTIVES

To predict kidney fibrosis in patients with chronic kidney disease using radiomics of two-dimensional ultrasound (B-mode) and Sound Touch Elastography (STE) images in combination with clinical features.

METHODS

The Mindray Resona 7 ultrasonic diagnostic apparatus with SC5-1U convex array probe (bandwidth frequency of 1-5 MHz) was used to perform two-dimensional ultrasound and STE software. The severity of cortical tubulointerstitial fibrosis was divided into three grades: mild interstitial fibrosis and tubular atrophy (IFTA), fibrotic area < 25%; moderate IFTA, fibrotic area 26-50%; and severe IFTA, fibrotic area > 50%. After extracting radiomics from B-mode and STE images in these patients, we analyzed two classification schemes: mild versus moderate-to-severe IFTA, and mild-to-moderate versus severe IFTA. A nomogram was constructed based on multiple logistic regression analyses, combining clinical and radiomics. The performance of the nomogram for differentiation was evaluated using receiver operating characteristic (ROC), calibration, and decision curves.

RESULTS

A total of 150 patients undergoing kidney biopsy were enrolled (mild IFTA: n = 74; moderate IFTA: n = 33; severe IFTA: n = 43) and randomized into training (n = 105) and validation cohorts (n = 45). To differentiate between mild and moderate-to-severe IFTA, a nomogram incorporating STE radiomics, albumin, and estimated glomerular filtration (eGFR) rate achieved an area under the ROC curve (AUC) of 0.91 (95% confidence interval [CI]: 0.85-0.97) and 0.85 (95% CI: 0.77-0.98) in the training and validation cohorts, respectively. Between mild-to-moderate and severe IFTA, the nomogram incorporating B-mode and STE radiomics features, age, and eGFR achieved an AUC of 0.93 (95% CI: 0.89-0.98) and 0.83 (95% CI: 0.70-0.95) in the training and validation cohorts, respectively. Finally, we performed a decision curve analysis and found that the nomogram using both radiomics and clinical features exhibited better predictability than any other model (DeLong test, p < 0.05 for the training and validation cohorts).

CONCLUSION

A nomogram based on two-dimensional ultrasound and STE radiomics and clinical features served as a non-invasive tool capable of differentiating kidney fibrosis of different severities.

KEY POINTS

• Radiomics calculated based on the ultrasound imaging may be used to predict the severities of kidney fibrosis. • Radiomics may be used to identify clinical features associated with the progression of tubulointerstitial fibrosis in patients with CKD. • Non-invasive ultrasound imaging-based radiomics method with accuracy aids in detecting renal fibrosis with different IFTA severities.

MRI texture-based machine learning models for the evaluation of renal function on different segmentations: a proof-of-concept study

BACKGROUND

To develop and validate an MRI texture-based machine learning model for the noninvasive assessment of renal function.

METHODS

A retrospective study of 174 diabetic patients (training cohort, n = 123; validation cohort, n = 51) who underwent renal MRI scans was included. They were assigned to normal function (n = 71), mild or moderate impairment (n = 69), and severe impairment groups (n = 34) according to renal function. Four methods of kidney segmentation on T2-weighted images (T2WI) were compared, including regions of interest covering all coronal slices (All-K), the largest coronal slices (LC-K), and subregions of the largest coronal slices (TLCO-K and PIZZA-K). The speeded-up robust features (SURF) and support vector machine (SVM) algorithms were used for texture feature extraction and model construction, respectively. Receiver operating characteristic (ROC) curve analysis was used to evaluate the diagnostic performance of models.

RESULTS

The models based on LC-K and All-K achieved the nonsignificantly highest accuracy in the classification of renal function (all p values > 0.05). The optimal model yielded high performance in classifying the normal function, mild or moderate impairment, and severe impairment, with an area under the curve of 0.938 (95% confidence interval [CI] 0.935-0.940), 0.919 (95%CI 0.916-0.922), and 0.959 (95%CI 0.956-0.962) in the training cohorts, respectively, as well as 0.802 (95%CI 0.800-0.807), 0.852 (95%CI 0.846-0.857), and 0.863 (95%CI 0.857-0.887) in the validation cohorts, respectively.

CONCLUSION

We developed and internally validated an MRI-based machine-learning model that can accurately evaluate renal function. Once externally validated, this model has the potential to facilitate the monitoring of patients with impaired renal function.

Machine learning-based multimodal MRI texture analysis for assessing renal function and fibrosis in diabetic nephropathy: a retrospective study

Introduction

Diabetic nephropathy (DN) has become a major public health burden in China. A more stable method is needed to reflect the different stages of renal function impairment. We aimed to determine the possible practicability of machine learning (ML)-based multimodal MRI texture analysis (mMRI-TA) for assessing renal function in DN.

Methods

For this retrospective study, 70 patients (between 1 January 2013 and 1 January 2020) were included and randomly assigned to the training cohort (n1 = 49) and the testing cohort (n2 = 21). According to the estimated glomerular filtration rate (eGFR), patients were assigned into the normal renal function (normal-RF) group, the non-severe renal function impairment (non-sRI) group, and the severe renal function impairment (sRI) group. Based on the largest coronal image of T2WI, the speeded up robust features (SURF) algorithm was used for texture feature extraction. Analysis of variance (ANOVA) and relief and recursive feature elimination (RFE) were applied to select the important features and then support vector machine (SVM), logistic regression (LR), and random forest (RF) algorithms were used for the model construction. The values of area under the curve (AUC) on receiver operating characteristic (ROC) curve analysis were used to assess their performance. The robust T2WI model was selected to construct a multimodal MRI model by combining the measured BOLD (blood oxygenation level-dependent) and diffusion-weighted imaging (DWI) values.

Results

The mMRI-TA model achieved robust and excellent performance in classifying the sRI group, non-sRI group, and normal-RF group, with an AUC of 0.978 (95% confidence interval [CI]: 0.963, 0.993), 0.852 (95% CI: 0.798, 0.902), and 0.972 (95% CI: 0.995, 1.000), respectively, in the training cohort and 0.961 (95% CI: 0.853, 1.000), 0.809 (95% CI: 0.600, 0.980), and 0.850 (95% CI: 0.638, 0.988), respectively, in the testing cohort.

Discussion

The model built from multimodal MRI on DN outperformed other models in assessing renal function and fibrosis. Compared to the single T2WI sequence, mMRI-TA can improve the performance in assessing renal function.

Radiomics-Based Assessment of OCT Angiography Images for Diabetic Retinopathy Diagnosis

Purpose

To evaluate the diagnostic accuracy of machine learning (ML) techniques applied to radiomic features extracted from OCT and OCT angiography (OCTA) images for diabetes mellitus (DM), diabetic retinopathy (DR), and referable DR (R-DR) diagnosis.

Design

Cross-sectional analysis of a retinal image dataset from a previous prospective OCTA study (ClinicalTrials.govNCT03422965).

Participants

Patients with type 1 DM and controls included in the progenitor study.

Methods

Radiomic features were extracted from fundus retinographies, OCT, and OCTA images in each study eye. Logistic regression, linear discriminant analysis, support vector classifier (SVC)-linear, SVC-radial basis function, and random forest models were created to evaluate their diagnostic accuracy for DM, DR, and R-DR diagnosis in all image types.

Main Outcome Measures

Area under the receiver operating characteristic curve (AUC) mean and standard deviation for each ML model and each individual and combined image types.

Results

A dataset of 726 eyes (439 individuals) were included. For DM diagnosis, the greatest AUC was observed for OCT (0.82, 0.03). For DR detection, the greatest AUC was observed for OCTA (0.77, 0.03), especially in the 3 × 3 mm superficial capillary plexus OCTA scan (0.76, 0.04). For R-DR diagnosis, the greatest AUC was observed for OCTA (0.87, 0.12) and the deep capillary plexus OCTA scan (0.86, 0.08). The addition of clinical variables (age, sex, etc.) improved most models AUC for DM, DR and R-DR diagnosis. The performance of the models was similar in unilateral and bilateral eyes image datasets.

Conclusions

Radiomics extracted from OCT and OCTA images allow identification of patients with DM, DR, and R-DR using standard ML classifiers. OCT was the best test for DM diagnosis, OCTA for DR and R-DR diagnosis and the addition of clinical variables improved most models. This pioneer study demonstrates that radiomics-based ML techniques applied to OCT and OCTA images may be an option for DR screening in patients with type 1 DM.

Financial Disclosures

Proprietary or commercial disclosure may be found after the references.

Radiomics: a primer on high-throughput image phenotyping

Radiomics is a high-throughput approach to image phenotyping. It uses computer algorithms to extract and analyze a large number of quantitative features from radiological images. These radiomic features collectively describe unique patterns that can serve as digital fingerprints of disease. They may also capture imaging characteristics that are difficult or impossible to characterize by the human eye. The rapid development of this field is motivated by systems biology, facilitated by data analytics, and powered by artificial intelligence. Here, as part of Abdominal Radiology's special issue on Quantitative Imaging, we provide an introduction to the field of radiomics. The technique is formally introduced as an advanced application of data analytics, with illustrating examples in abdominal radiology. Artificial intelligence is then presented as the main driving force of radiomics, and common techniques are defined and briefly compared. The complete step-by-step process of radiomic phenotyping is then broken down into five key phases. Potential pitfalls of each phase are highlighted, and recommendations are provided to reduce sources of variation, non-reproducibility, and error associated with radiomics.

Ultrasound Based Radiomics Features of Chronic Kidney Disease

RATIONALE AND OBJECTIVES

Interstitial fibrosis, common to most chronic kidney diseases, can potentially affect the speckle patterns of kidney ultrasound (US). Here we use Radiomics features derived from US images to identify kidneys with chronic kidney disease.

MATERIALS AND METHODS

B-mode US without speckle reduction was performed on a cohort of CKD patients (n = 75) and healthy subjects (n = 27). Images of the patients with renal cysts, agenesis and calculi were excluded. After background subtraction, regions of interest were selected from each kidney. Four hundred and sixty-five Radiomics features including first and second-order gray level statistics were calculated on the selected regions. Second-order features were also calculated on wavelet transformed images. A random forest model was used to identify the most important features that can differentiate healthy and diseased kidneys. The ten most important features, based on the Gini index, were used to train a support vector machine. Synthetic minority oversampling technique was used to remove over fitting.

RESULTS

Wavelet transformed, Gray Level Run Length Matrix based Normalized Run Length Non-uniformity, WT (LH) (GRLN) was identified as the most significant feature in differentiating CKD and healthy kidneys (accuracy - 0.85, sensitivity - 1.0). The mean WT (LH) GRLN of healthy kidneys (0.40 ± 0.01) was significantly higher (p < 0.01) than that of the CKD kidneys (0.24 ± 0.01). According to the Gini Index, the differentiability of WT (LH) GRLN was highest when the long axis of the kidney was oriented perpendicular to the columns of the image matrix.

CONCLUSION

Radiomics features based on wavelet transformation are sensitive to directionality of US speckle patters and can be successfully used to differentiate CKD and healthy US kidney images.

Single-Photon Emission Computed Tomography/Computed Tomography Image-Based Radiomics for Discriminating Vertebral Bone Metastases From Benign Bone Lesions in Patients With Tumors

Purpose: The purpose of this study was to investigate the feasibility of Single-Photon Emission Computed Tomography/Computed Tomography (SPECT/CT) image-based radiomics in differentiating bone metastases from benign bone lesions in patients with tumors. Methods: A total of 192 lesions from 132 patients (134 in the training group, 58 in the validation group) diagnosed with vertebral bone metastases or benign bone lesions were enrolled. All images were evaluated and diagnosed independently by two physicians with more than 20 years of diagnostic experience for qualitative classification, the images were imported into MaZda software in Bitmap (BMP) format for feature extraction. All radiomics features were selected by least absolute shrinkage and selection operator (LASSO) regression and 10-fold cross-validation algorithms after the process of normalization and correlation analysis. Based on these selected features, two models were established: The CT model and SPECT model (radiomics features were derived from CT and SPECT images, respectively). In addition, a combination model (ComModel) combined CT and SPECT features was developed in order to better evaluate the predictive performance of radiomics models. Subsequently, the diagnostic performance between each model was separately evaluated by a confusion matrix. Results: There were 12, 13, and 18 features contained within the CT, SPECT, and ComModel, respectively. The constructed radiomics models based on SPECT/CT images to discriminate between bone metastases and benign bone lesions not only had high diagnostic efficacy in the training group (AUC of 0.894, 0.914, 0.951 for CT model, SPECT model, and ComModel, respectively), but also performed well in the validation group (AUC; 0.844, 0.871, 0.926). The AUC value of the human experts was 0.849 and 0.839 in the training and validation groups, respectively. Furthermore, both SPECT model and ComModel show higher classification performance than human experts in the training group (P = 0.021 and P = 0.001, respectively) and the validation group (P = 0.037 and P = 0.007, respectively). All models showed better diagnostic accuracy than human experts in the training group and the validation group. Conclusion: Radiomics derived from SPECT/CT images could effectively discriminate between bone metastases and benign bone lesions. This technique may be a new non-invasive way to help prevent unnecessary delays in diagnosis and a potential contribution in disease staging and treatment planning.

Capability of intravoxel incoherent motion and diffusion tensor imaging to detect early kidney injury in type 2 diabetes

OBJECTIVES

To prospectively investigate the capability of intravoxel incoherent motion (IVIM) and conventional diffusion tensor imaging (DTI) to identify early kidney function injury in type 2 diabetes.

METHODS

Forty-one diabetes patients (normoalbuminuria: n = 27; microalbuminuria: n = 14) and 28 volunteers were recruited. All participants were examined using DTI and IVIM with 3.0-T MRI. DTI parameters (mean diffusivity [MD], fractional anisotropy [FA]), and IVIM parameters (true diffusion coefficient [D], pseudo-diffusion coefficient [D*], and pseudo-diffusion component fraction [f]) were measured in the renal parenchyma (cortex and medulla) by two experienced radiologists independently. Image features were compared among the groups using separate one-way analyses of variance. Diagnostic performances of various diffusion parameters for predicting diabetic renal damage were compared.

RESULTS

The medullary D and FA values were significantly different among the microalbuminuria subgroup, normoalbuminuria subgroup, and control group (all p < 0.001). In medulla, area under the curve (AUC) values for combined FA and D were significantly higher than single FA (AUC = 0.938, 0.769, respectively; p = 0.003), and the combined AUC of FA and D was numerically higher than that of single D (0.938 vs 0.878, p > 0.05). AUC of combined FA and D was 0.985, not significantly different from individual AUC for FA and D (AUC = 0.909 and 0.952, respectively; all p > 0.05) in differentiating the microalbuminuria subgroup from the control group.

CONCLUSION

IVIM-derived D and DTI-derived FA values were better than other parameters for evaluating early kidney impairment of diabetes. The single indicator FA and D performed as well as the combined diagnostic indicator in the medulla for differentiating the microalbuminuria subgroup from the control group.

KEY POINTS

• We speculated that early renal progression in type 2 diabetes result from restricted tubular flow and kidney tubule dysregulation may precede or at least accompany abnormal glomerular changes. • In medulla, the AUC values of FA and D and the combination of FA and D obtained by comparing the microalbuminuria subgroup with the control group were 0.909, 0.952, and 0.985, respectively. • IVIM-derived D and DTI-derived FA are effective MR biomarkers to evaluate early alterations of the renal function in patients with diabetes.

Application of MR Imaging Features in Differentiation of Renal Changes in Patients With Stage III Type 2 Diabetic Nephropathy and Normal Subjects

OBJECTIVE

The objective of the study was to explore the value of MRI texture features based on T1WI, T2-FS and diffusion-weighted imaging (DWI) in differentiation of renal changes in patients with stage III type 2 diabetic nephropathy (DN) and normal subjects.

MATERIALS AND METHODS

A retrospective analysis was performed to analyze 44 healthy volunteers (group A) and 40 patients with stage III type 2 diabetic nephropathy (group B) with microalbuminuria. Urinary albumin to creatinine ratio (ACR) <30 mg/g, estimated glomerular filtration rate (eGFR) in the range of 60-120 ml/(min 1.73 m²), and randomly divided into primary cohort and test cohort. Conventional MRI and DWI of kidney were performed using 1.5 T magnetic resonance imaging (MRI). The outline of the renal parenchyma was manually labeled in fat-suppressed T2-weighted imaging (FS-T2WI), and PyRadiomics was used to extract radiomics features. The radiomics features were then selected by the least absolute shrinkage and selection operator (LASSO) method.

RESULTS

There was a significant difference in sex and body mass index (BMI) (P <0.05) in the primary cohort, with no significant difference in age. In the final results, the wavelet and Laplacian-Gaussian filtering are used to extract 1,892 image features from the original T1WI image, and the LASSO algorithm is used for selection. One first-order feature and six texture features are selected through 10 cross-validations. In the mass, 1,638 imaging extracts features from the original T2WI image.1 first-order feature and 5 texture features were selected. A total of 1,241 imaging features were extracted from the original ADC images, and 5 texture features were selected. Using LASSO-Logistic regression analysis, 10 features were selected for modeling, and a combined diagnosis model of diabetic nephropathy based on texture features was established. The average unit cost in the logistic regression model was 0.98, the 95% confidence interval for the predictive efficacy was 0.9486-1.0, specificity 0.97 and precision 0.93, particularly. ROC curves also revealed that the model could distinguish with high sensitivity of at least 92%.

CONCLUSION

In consequence, the texture features based on MR have broad application prospects in the early detection of DN as a relatively simple and noninvasive tool without contrast media administration.

Diffusion Tensor Imaging in Rat Models of Preclinical Diabetic Nephropathy: A Preliminary Study

PURPOSE

This study aimed to investigate the value of diffusion tensor imaging to assess renal injury in a rat model of preclinical diabetic nephropathy.

METHODS

Twenty-eight male Sprague Dawley rats were divided into two groups: the normal control (NC) group of 10 rats and the diabetic nephropathy (DN) group of 18 rats. Eight weeks after diabetes induction by streptozotocin, 3.0-T magnetic resonance (MR) imaging (b = 0 and 600 s/mm², 15 diffusion directions) using a 32-channel knee coil was performed. After MR imaging, we measured serum creatinine, and collected double kidney tissues for pathology. The apparent diffusion coefficients(ADC) and fractional anisotropy(FA) values of the renal cortex and medulla were calculated for all kidneys. Physiological parameters, laboratory parameters, and imaging results were compared between the two groups.

RESULTS

All DN group animals developed hyperglycemia, polyuria, and emaciation. Serum creatinine was not significantly different between the groups (P > 0.05). Urinary albumin at 2, 4, and 8 weeks was higher in the DN group than in the NC group but <20 µg/min (P < 0.05). Pathologically, renal damage in the DN rats was observed. The ADC value was significantly increased in DN animals in the cortex (1.75×10^-3mm²/s),medulla(1.53×10^-3mm²/s)compared with NC group(cortex, 1.52×10^-3mm²/s; medulla,1.35×10^-3mm²/s). The FA value was significantly reduced in DN animals in the cortex (0.21),medulla(0.25)compared with NC group(cortex,0.26;medulla,0.3).

CONCLUSIONS

Increased apparent diffusion coefficients and decreased fractional anisotropy values on diffusion tensor imaging were associated with preclinical DN. Diffusion tensor imaging may be useful in early, non-invasive, quantitative detection, and therapy monitoring of DN.