Explore the Value of Multi-Parameter MRI in Non-Invasive Assessment of Prognostic Risk and Oxford Classification in Children with IgAN or IgAVN.

PURPOSE: To explore the Oxford classification and prognostic risk stratification of the non-invasive evaluation of immunoglobulin A nephropathy (IgAN) or immunoglobulin A vasculitis with nephritis (IgAVN) in children using multiparametric magnetic resonance imaging (MRI). MATERIALS AND METHODS: Forty-four children diagnosed with IgAN or IgAVN were included. Patients with 80-month risk scores >10% were categorized as the high-risk group, while others constituted the low-risk group. The T2* and apparent diffusion coefficient (ADC) values of the renal cortex and medulla were measured. Clinical and pathological parameters were also assessed. Univariate and multivariate logistic regression analyses were performed to identify the indicators associated with the high-risk group. Receiver operating characteristic (ROC) curves were drawn and the areas under the curve (AUCs) were calculated to evaluate the diagnostic performance variables for differentiating the high-risk group from the low-risk group. RESULTS: Only the T2*Cortex and mean arterial pressure (MAP) were independently reliable in both the univariate and multivariate analyses. The AUCs for differentiating the high-risk group from the low-risk group of T2*Cortex, MAP, and their combination model were 0.907, 0.881, and 0.947, respectively. CONCLUSIONS: Multiparametric MRI parameters, especially T2* values, could be used as new biomarkers to provide a new dimension in chronic kidney disease-related research and could play an important role in the non-invasive prognosis of children with IgAN or IgAVN.

Identification of Calculous Pyonephrosis by CT-Based Radiomics and Deep Learning.

Urgent detection of calculous pyonephrosis is crucial for surgical planning and preventing severe outcomes. This study aims to evaluate the performance of computed tomography (CT)-based radiomics and a three-dimensional convolutional neural network (3D-CNN) model, integrated with independent clinical factors, to identify patients with calculous pyonephrosis. We recruited 182 patients receiving either percutaneous nephrostomy tube placement or percutaneous nephrolithotomy for calculous hydronephrosis or pyonephrosis. The regions of interest were manually delineated on plain CT images and the CT attenuation value (HU) was measured. Radiomics analysis was performed using least absolute shrinkage and selection operator (LASSO). A 3D-CNN model was also developed. The better-performing machine-learning model was combined with independent clinical factors to build a comprehensive clinical machine-learning model. The performance of these models was assessed using receiver operating characteristic analysis and decision curve analysis. Fever, blood neutrophils, and urine leukocytes were independent risk factors for pyonephrosis. The radiomics model showed higher area under the curve (AUC) than the 3D-CNN model and HU (0.876 vs. 0.599, 0.578; p = 0.003, 0.002) in the testing cohort. The clinical machine-learning model surpassed the clinical model in both the training (0.975 vs. 0.904, p = 0.019) and testing (0.967 vs. 0.889, p = 0.045) cohorts.

Noninvasive assessment of renal function and fibrosis in CKD patients using histogram analysis based on diffusion kurtosis imaging.

PURPOSE: To investigate the potential of histogram analysis based on diffusion kurtosis imaging (DKI) in evaluating renal function and fibrosis associated with chronic kidney disease (CKD). MATERIALS AND METHODS: Thirty-six CKD patients were enrolled, and DKI was performed in all patients before the renal biopsy. The histogram parameters of diffusivity (D) and kurtosis (K) were obtained using FireVoxel. The histogram parameters between the stable [estimated glomerular filtration rate (eGFR) ≥ 60 ml/min/1.73 m2] and impaired (eGFR < 60 ml/min/1.73 m2) eGFR group were compared. Besides, patients were classified into mild, moderate, and severe fibrosis group using a semi-quantitative standard. The correlations of histogram parameters with eGFR and fibrosis scores were investigated and the diagnostic performances of histogram parameters in assessing renal dysfunction and fibrosis were analyzed. The added value of combination of most significant parameter with 24 h urinary protein (24 h-UPRO) in evaluating fibrosis was also explored. RESULTS: Seven D histogram parameters in cortex (mean, median, 10th, 25th, 75th, 90th percentiles and entropy), two D histogram parameters in medulla (75th, 90th percentiles), seven K histogram parameters in cortex (mean, min, median, 10th, 25th, 75th, 90th percentiles) and three K histogram parameters in medulla (mean, median, 25th percentile) were significantly different between the two groups. The Dmean of cortex was the most relevant parameter to eGFR (r = 0.648, P < 0.001) and had the largest area under the curve (AUC) for differentiating the stable from impaired eGFR group [AUC = 0.889; 95% confidence interval (CI) 0.728-0.970]. The K90th of cortex presented the strongest correlation with fibrosis scores (r = 0.575, P < 0.001) and achieved the largest AUC for distinguishing the mild from moderate to severe fibrosis group (AUC = 0.849, 95% CI 0.706-0.993). Combining the K90th in cortex with 24 h-UPRO gained statistically higher AUC value (AUC = 0.880, 95% CI 0.763-0.996). CONCLUSION: Histogram analysis based on DKI is practicable for the noninvasive assessment of renal function and fibrosis in CKD patients.

Non-invasive evaluation of the pathological and functional characteristics of chronic kidney disease by diffusion kurtosis imaging and intravoxel incoherent motion imaging: comparison with conventional DWI.

OBJECTIVE: To explore the diagnostic performance of diffusion kurtosis imaging (DKI) and incoherent intravoxel movement (IVIM) in evaluating the clinical and pathological characteristics in chronic kidney disease (CKD) compared to conventional diffusion-weighted imaging (DWI). METHODS: Forty-nine CKD patients and 24 healthy volunteers were included in this retrospective study from September 2020 to September 2021. All participants underwent MRI examinations before percutaneous renal biopsy. Coronal T2WI, axial T1WI and T2WI, and DWI (including IVIM and DKI) sequences obtained in one scan. We measured the apparent diffusion coefficient (ADC), true diffusion coefficient (Dt), pseudo-diffusion coefficient (Dp), perfusion fraction (fp), mean kurtosis (MK), and mean diffusivity (MD) values. One-way analysis of variance, correlation analysis, and receiver operating characteristic curve analysis were used in our study. RESULTS: Cortex and medulla ADC, MK, Dt, fp were significantly different between the healthy volunteers and CKD stages 1-2 (all p < 0.05). All diffusion parameters showed significant differences between CKD stages 1-2 and CKD stages 3-5 (all p < 0.05). Except for the uncorrelation between MDMedulla and vascular lesion score, all other diffusion parameters were low-to-moderately related to clinical and pathological indicators. fpMedulla was the best parameter to differentiate healthy volunteers from CKD stages 1-2. MKCortex was the best parameter to differentiate CKD stages 1-2 from that CKD stages 3-5. CONCLUSION: Renal cortex and medulla fp, Dt, and MK can provide more valuable information than ADC values for the evaluation of clinical and pathological characteristics of CKD patients, and thus can provide auxiliary diagnosis for fibrosis assessment and clinical management of CKD patients. ADVANCES IN KNOWLEDGE: IVIM and DKI can provide more diagnostic valuable information for CKD patients than conventional DWI.

Noninvasive Assessment of the Renal Function, Oxford Classification and Prognostic Risk Stratification of IgAN by Using Intravoxel Incoherent Motion Diffusion-Weighted Imaging and Blood Oxygenation Level-Dependent MRI.

BACKGROUND: Immunoglobulin A nephropathy (IgAN) is the most common primary glomerulonephritis worldwide. Oxford classification including mesangial hypercellularity (M), endothelial hypercellularity (E), segmental sclerosis (S), interstitial fibrosis/tubular atrophy (T), and crescent (C) were recommended to predict the prognosis of IgAN. PURPOSE: To explore whether multiparametric magnetic resonance imaging (MRI) can be applied to assess the renal function, Oxford classification, and risk of progression to end-stage kidney disease within 5 years of IgAN. STUDY TYPE: Prospective. POPULATION: A total of 46 patients with pathologically confirmed IgAN and 20 healthy volunteers. FIELD STRENGTH/SEQUENCE: A 3-T, blood oxygenation level-dependent (BOLD)-MRI, intravoxel incoherent motion diffusion-weighted imaging (IVIM-DWI). ASSESSMENT: Two radiologists measured the cortex and medulla T2*, apparent diffusion coefficient (ADC), true diffusion (Dt), pseudo-diffusion (Dp), perfusion fraction (fp). All participants were divided into three groups: group 1, healthy volunteers; group 2, patients with estimated glomerular filtration rate (eGFR) ≥60 mL/min/1.73 m2 ; group 3, patients with eGFR <60 mL/min/1.73 m2 . Or two groups: group A, 5-year risk scores ≤10% and group B, 5-year risk scores >10%. STATISTICAL TESTS: Intraclass correlation coefficient, one-way analysis of variance, least-significant difference, Student's t-test, Pearson product-moment correlation, Spearman's rank correlation, and receiver operating characteristics (ROC) with the area under the curve (AUC). A P value <0.05 was considered statistically significant. RESULTS: Except for cortical Dp, all other MRI parameters showed significant differences between group 1 and group 2. None of the MRI parameters showed a significant correlation with M, E, or C scores. Cortical T2*, Dt, fp, and medullary Dt and fp showed low-to-moderate significant correlations with S scores. Except for cortical and medullary Dp, all other MRI parameters were significantly correlated with T scores. Cortical Dt showed the largest AUC for differentiating group A from group B (AUC = 0.927) and T0 from T1/T2 (AUC = 0.963). DATA CONCLUSION: Imaging by IVIM-DWI and BOLD-MRI could facilitate noninvasive assessment of the renal function, Oxford classification, and prognostic risk of IgAN patients. EVIDENCE LEVEL: 2. TECHNICAL EFFICACY: Stage 3.

Noninvasive assessment of clinical and pathological characteristics of patients with IgA nephropathy by diffusion kurtosis imaging.

OBJECTIVES: To explore the diagnostic performance of diffusion kurtosis imaging (DKI) in evaluating the clinical and pathological characteristics of patients with immunoglobulin A nephropathy (IgAN) compared with conventional DWI. MATERIALS AND METHODS: A total of 28 IgAN patients and 14 healthy volunteers prospectively underwent MRI examinations including coronal T2WI, axial T1WI, T2WI, and DWI sequences from September 2020 to August 2021. We measured mean kurtosis (MK), mean diffusivity (MD), and apparent diffusion coefficient (ADC) by using MR Body Diffusion Toolbox v1.4.0 (Siemens Healthcare). Patients were divided into three groups according to their estimated glomerular filtration rate (eGFR) (Group1, healthy volunteers without kidney disease or other diseases that affect renal function; Group2, IgAN patients with eGFR > 60 mL/min/1.73 m2; Group3, IgAN patients with eGFR < 60 mL/min/1.73 m2). One-way analysis of variance, Pearson or Spearman correlation, and receiver operating characteristic curves were applied in our statistical analysis. RESULTS: MKCortex and ADCCortex showed significant differences between the Group1 and Group2. MKCortex, MDCortex, ADCCortex, MKMedulla, and ADCMedulla showed significant differences between Group2 and Group3. MKCortex had the highest correlation with CKD stages (r = 0.749, p < 0.001), and tubulointerstitial lesion score (r = 0.656, p < 0.001). MDCortex had the highest correlation with glomerular lesion score (r = - 0.475, p = 0.011). MKCortex had the highest AUC (AUC = 0.923) for differentiating Group1 from Group2, and MDCortex had the highest AUC (AUC = 0.924) for differentiating Group2 from Group3, followed by MKMedulla (AUC = 0.923). CONCLUSIONS: DKI is a feasible and reliable technique that can assess the clinical and pathological characteristics of IgAN patients and can provide more valuable information than conventional DWI, especially MKCortex.

Erratum: Author Correction: Advancing COVID-19 diagnosis with privacy-preserving collaboration in artificial intelligence.

[This corrects the article DOI: 10.1038/s42256-021-00421-z.].

Assessment of renal function using magnetic resonance quantitative histogram analysis based on spatial labeling with multiple inversion pulses.

BACKGROUND: The incidence of chronic kidney disease (CKD) is high, and is easy to develop into end-stage renal disease (ESRD), which requires kidney dialysis or kidney transplantation. Therefore, we want to explore the clinical value of magnetic resonance quantitative histogram analysis based on spatial labeling with multiple inversion pulses (SLEEK) in assessing renal function in the early stage. METHODS: One hundred and twenty-nine patients underwent abdominal MRI examination, including a coronal SLEEK sequence. The patients were divided into the control group [CG, 47 cases, estimated glomerular filtration rate (eGFR) >90], the mild renal function impairment (mRI) group (48 cases, eGFR =60-90), and the moderate to severe renal function impairment (m-sRI) group (34 cases, eGFR <60). Two experienced radiologists delineated cortex and medulla regions of interest (ROIs) on SLEEK images to obtain cortex and medulla quantitative histogram parameters [Mean, Median, Percentiles (5th, 10th, 25th, 75th, and 90th), Skewness, Kurtosis, and Entropy] using FireVoxel. These histogram parameters were compared by proper statistical methods such as one-way analysis of variance, the χ2 test, and receiver operating characteristic (ROC) curve analysis. RESULTS: Four histogram parameters (Inhomogeneitycortex, Skewnesscortex, Kurtosismedulla, and Entropymedulla) differed significantly between the CG and the mRI group. One medulla (Entropymedulla) and nine cortex (Meancortex, Mediancortex, Kurtosiscortex, Entropycortex, and 5th, 10th, 25th, 75th, and 90th Percentilecortex) histogram parameters were significantly different between the m-RI and m-sRI groups. The most relevant parameter to eGFR was Inhomogenitycortex (r=-0.450, P<0.001). Inhomogeneitycortex had the largest area under the curve (AUC) for differentiating the mRI group from the CG (AUC =0.718; 95% CI: 0.616-0.806), while 25th Percentilecortex generated the largest AUC (AUC =0.786; 95% CI: 0.681-0.869) for differentiating the mRI and m-sRI groups. CONCLUSIONS: Quantitative histogram parameters based on a SLEEK sequence can be used to supplement renal dysfunction assessment. Cortex histogram parameters are more valuable for evaluating renal function than medulla histogram parameters.

Advancing COVID-19 Diagnosis with Privacy-Preserving Collaboration in Artificial Intelligence.

Artificial intelligence (AI) provides a promising substitution for streamlining COVID-19 diagnoses. However, concerns surrounding security and trustworthiness impede the collection of large-scale representative medical data, posing a considerable challenge for training a well-generalised model in clinical practices. To address this, we launch the Unified CT-COVID AI Diagnostic Initiative (UCADI), where the AI model can be distributedly trained and independently executed at each host institution under a federated learning framework (FL) without data sharing. Here we show that our FL model outperformed all the local models by a large yield (test sensitivity /specificity in China: 0.973/0.951, in the UK: 0.730/0.942), achieving comparable performance with a panel of professional radiologists. We further evaluated the model on the hold-out (collected from another two hospitals leaving out the FL) and heterogeneous (acquired with contrast materials) data, provided visual explanations for decisions made by the model, and analysed the trade-offs between the model performance and the communication costs in the federated training process. Our study is based on 9,573 chest computed tomography scans (CTs) from 3,336 patients collected from 23 hospitals located in China and the UK. Collectively, our work advanced the prospects of utilising federated learning for privacy-preserving AI in digital health.

Noninvasive assessment of kidney dysfunction in children by using blood oxygenation level-dependent MRI and intravoxel incoherent motion diffusion-weighted imaging.

OBJECTIVES: To explore whether multiparametric approach including blood oxygenation level-dependent MRI (BOLD-MRI) and intravoxel incoherent motion diffusion-weighted imaging (IVIM-DWI) can be applied in the assessment of renal function in children with chronic kidney disease (CKD). MATERIALS AND METHODS: This prospective study included 74 children (CKD stage 1-3, 51; CKD stage 4-5, 12; healthy volunteers, 11) for renal MRI examinations including coronal T2WI, axial T1WI and T2WI, BOLD-MRI, and DWI sequences. We measured the renal cortex and medulla T2*, ADC, Dt, Dp, and fp values on BOLD and DWI images. Appropriate statistical methods were applied for comparing MRI-derived parameters among the three groups and calculating the correlation coefficients between MRI-derived parameters and clinical data. Receiver operating characteristic (ROC) curves were used to assess the diagnostic performance of MRI-derived parameters. RESULTS: There were significant differences in cortex T2*, ADC, Dt, fp and medulla T2*, ADC, Dt among the three groups. Cortex T2*, ADC, Dt, fp and medulla T2*, ADC, Dt had a trend: CKD stage 4-5 < CKD stage 1-3 < healthy volunteers. Cortex and medulla T2*, ADC, Dt were significantly correlated with eGFR, serum creatinine (Scr), cystatin C. In addition, cortex T2* and eGFR showed the highest correlation coefficient (r = 0.824, p < 0.001). Cortex Dt and medulla T2* were optimal parameters for differentiating healthy volunteers and CKD stage 1-3 or CKD stage 4-5 and CKD stage 1-3, respectively. CONCLUSIONS: BOLD-MRI and IVIM-DWI might be used as a feasible method for noninvasive assessment of renal function in children with CKD.

Characteristics of COVID-19 patients with preexisting CKD history.

PURPOSE: This paper was intended to describe the characteristics of coronavirus disease 2019 (COVID-19) patients with known chronic kidney disease (CKD) history. METHODS: Clinical information of 20 COVID-19 pneumonia patients with CKD history diagnosed between January 20th and March 1st, 2020 were collected in Tongji Hospital, Wuhan. We listed the clinical baseline data, laboratory findings, chest computed tomography (CT) changes and processed a short period of follow-up of these 20 patients. RESULTS: Based on the estimated glomerular filtration rate (eGFR) on admission, 6 patients were classified as stage 2 of CKD, 5 were as 3a, 2 were as 3b, 3 were as 4 and 4 were as 5, respectively. COVID-19 patients with CKD history were elder and hypertension was the most common comorbidity. Cough and fever accounted for more than 80% of the infectious cases. Lymphopenia, increased D-dimer and elevated infectious indications such as hypersensitive C response protein (hsCRP), interleukin-6 (IL-6) and tumor necrosis factor alpha (TNF-α) were also common among these patients. Ground-glass opacity (GGO) and consolidation were the major manifestations in CT scans. 4 patients died and 7 patients underwent acute kidney injury (AKI) during observation. Among 16 discharged patients, 12 were with stable renal function and 4 had deteriorating renal function compared with that of admission. CONCLUSION: Compared to general population infected with SARS-CoV-2, COVID-19 patients with CKD history had a preference to develop to severity with higher fatality rate.

One-stop assessment of renal function and renal artery in hypertensive patients with suspected renal dysfunction: non-enhanced MRI using spatial labeling with multiple inversion pulses.

OBJECTIVES: To assess whether spatial labeling with multiple inversion pulses (SLEEK) sequence can be employed as a one-stop assessment method for evaluating renal function and displaying renal artery in hypertensive patients with suspected renal dysfunction. METHODS: A total of 78 patients with suspected hypertensive renal damage were enrolled in this retrospective study. All patients underwent MRI examinations, and both SLEEK and DWI sequences were performed simultaneously. According to estimated glomerular filtration rate (eGFR), patients were divided into three groups (Group 1, eGFR> 90; Group 2, eGFR = 60-90; Group 3, eGFR< 60). Twenty-two of these patients also underwent CT angiography (CTA) examination. Comparison between CTA, DWI, and eGFR was performed to assess the value of SLEEK in evaluating renal function and displaying renal artery. RESULTS: The performance of SLEEK to display renal artery was highly consistent with the results of CTA (kappa = 0.713). The corticomedullary contrast ratio positively correlated with eGFR (p = 0.004, r = 0.322) and was significantly higher in SLEEK images than in DWI images in all three groups (p < 0.001). There was no significant difference in corticomedullary contrast ratio in SLEEK images between Group 1 and Group 2 (p = 0.285). However, the minimal renal cortical thickness, which significantly correlated with eGFR (p < 0.001, r = 0.866), was significantly different between Group 1 and Group 2 (p < 0.001). ROC analysis showed good diagnostic performance when differentiating patients with eGFR> 60 from those with eGFR< 60. CONCLUSIONS: The SLEEK sequence could evaluate simultaneously renal function through corticomedullary differentiation and renal arteries, enabling one-stop assessment in hypertensive patients with suspected renal dysfunction. KEY POINTS: • Spatial labeling with multiple inversion pulses (SLEEK) improves renal corticomedullary differentiation in hypertensive patients with renal dysfunction compared with DWI. • SLEEK clearly displays renal artery in hypertensive patients with renal dysfunction. • SLEEK could be utilized as a one-stop assessment method for evaluating renal function and renal artery in hypertensive patients.

Prediction of the World Health Organization Grade of rectal neuroendocrine tumors based on CT histogram analysis.

OBJECTIVES: To investigate the diagnostic value of contrast-enhanced computed tomography (CECT) histogram analysis in predicting the World Health Organization (WHO) grade of rectal neuroendocrine tumors (R-NETs). MATERIALS AND METHODS: A total of 61 (35 G1, 12 G2, 10 G3, and 4 NECs) patients who underwent preoperative CECT and treated with surgery to be confirmed as R-NETs were included in this study from January 2014 to May 2019. We depicted ROIs and measured the CECT texture parameters (mean, median, 10th, 25th, 75th, 90th percentiles, skewness, kurtosis, and entropy) from arterial phase (AP) and venous phase (VP) images by two radiologists. We calculated intraclass correlation coefficient (ICC) and compared the histogram parameters between low-grade (G1) and higher grade (HG) (G2/G3/NECs) by applying appropriate statistical method. We obtained the optimal parameters to identify G1 from HG using receiver operating characteristic (ROC) curves. RESULTS: The capability of AP and VP histogram parameters for differentiating G1 from HG was similar in several histogram parameters (mean, median, 10th, 25th, 75th, and 90th percentiles) (all p < 0.001). Skewness, kurtosis, and entropy on AP images showed no significant differences between G1 and HG (p = 0.853, 0.512, 0.557, respectively). Entropy on VP images was significantly different (p = 0.017) between G1 and HG, however, skewness and kurtosis showed no significant differences (p = 0.654, 0.172, respectively). ROC analysis showed a good predictive performance between G1 and HG, and the 75th (AP) generated the highest area under the curve (AUC = 0.871), followed by the 25th (AP), mean (VP), and median (VP) (AUC = 0.864). Combined the size of tumor and the 75th (AP) generated the highest AUC. CONCLUSIONS: CECT histogram parameters, including arterial and venous phases, can be used as excellent indicators for predicting G1 and HG of rectal neuroendocrine tumors, and the size of the tumor is also an important independent predictor.

Advancing COVID-19 diagnosis with privacy-preserving collaboration in artificial intelligence.

Artificial intelligence provides a promising solution for streamlining COVID-19 diagnoses; however, concerns surrounding security and trustworthiness impede the collection of large-scale representative medical data, posing a considerable challenge for training a well-generalized model in clinical practices. To address this, we launch the Unified CT-COVID AI Diagnostic Initiative (UCADI), where the artificial intelligence (AI) model can be distributedly trained and independently executed at each host institution under a federated learning framework without data sharing. Here we show that our federated learning framework model considerably outperformed all of the local models (with a test sensitivity/specificity of 0.973/0.951 in China and 0.730/0.942 in the United Kingdom), achieving comparable performance with a panel of professional radiologists. We further evaluated the model on the hold-out (collected from another two hospitals without the federated learning framework) and heterogeneous (acquired with contrast materials) data, provided visual explanations for decisions made by the model, and analysed the trade-offs between the model performance and the communication costs in the federated training process. Our study is based on 9,573 chest computed tomography scans from 3,336 patients collected from 23 hospitals located in China and the United Kingdom. Collectively, our work advanced the prospects of utilizing federated learning for privacy-preserving AI in digital health.

Gallbladder carcinoma: an initial clinical experience of reduced field-of-view diffusion-weighted MRI.

BACKGROUND: The purpose of this study is to compare the diagnostic value, imaging quality and apparent diffusion coefficient (ADC) value of reduced field-of-view diffusion-weight imaging (r-FOV DWI) and full field-of-view diffusion-weight imaging (f-FOV DWI) in patients with gallbladder carcinoma and other lesions of gallbladder. METHODS: Two hundred ninety-six patients with gallbladder diseases underwent both r-FOV DWI and f-FOV DWI on a 3.0 T MRI scanner. Two radiologists assessed subjective image quality parameters independently. The Wilcoxon signed-rank test was used to compare subjective qualitative image score. Objective quality values and the mean ADC values were analyzed by paired t-test. The correlation between pathological results and mean ADC value were estimated using Spearman rank correlation analysis. RESULTS: The CNR value (10.23 ± 2.92) and image quality score (13.84 ± 1.07) of r-FOV DWI were significantly higher than those of f-FOV DWI (5.24 ± 1.29 P<0.001; 10.41 ± 1.11 P<0.001). There was no significant difference between mean ADC values of the two DWI sequences for all three groups (Group 1, chronic cholecystitis; Group 2, benign lesions of gallbladder; Group 3, gallbladder carcinoma. P = 0.239, 0.974 and 0.226 respectively). For both DWI sequences, the mean ADC values were the highest in the group of cholecystitis and the lowest in the group of gallbladder carcinoma (2.49 ± 0.14 vs 1.49 ± 0.12; 2.50 ± 0.14 vs 1.50 ± 0.13, for f-FOV and r-FOV respectively), the differences among groups were statistically significant (P<0.01). The mean ADC values for both DWI sequences were negatively correlated with the group number, which increased with the malignant tendency of lesions (r = - 0.892, P<0.01; r = - 0.913, P<0.01 for f-FOV and r-FOV respectively). CONCLUSION: Reduced Field-of-view Diffusion-weighted MRI is a good tool to diagnosis the gallbladder carcinoma, with better image quality and without affecting ADC values.

A collaborative online AI engine for CT-based COVID-19 diagnosis.

Artificial intelligence can potentially provide a substantial role in streamlining chest computed tomography (CT) diagnosis of COVID-19 patients. However, several critical hurdles have impeded the development of robust AI model, which include deficiency, isolation, and heterogeneity of CT data generated from diverse institutions. These bring about lack of generalization of AI model and therefore prevent it from applications in clinical practices. To overcome this, we proposed a federated learning-based Unified CT-COVID AI Diagnostic Initiative (UCADI, http://www.ai-ct-covid.team/), a decentralized architecture where the AI model is distributed to and executed at each host institution with the data sources or client ends for training and inferencing without sharing individual patient data. Specifically, we firstly developed an initial AI CT model based on data collected from three Tongji hospitals in Wuhan. After model evaluation, we found that the initial model can identify COVID from Tongji CT test data at near radiologist-level (97.5% sensitivity) but performed worse when it was tested on COVID cases from Wuhan Union Hospital (72% sensitivity), indicating a lack of model generalization. Next, we used the publicly available UCADI framework to build a federated model which integrated COVID CT cases from the Tongji hospitals and Wuhan Union hospital (WU) without transferring the WU data. The federated model not only performed similarly on Tongji test data but improved the detection sensitivity (98%) on WU test cases. The UCADI framework will allow participants worldwide to use and contribute to the model, to deliver a real-world, globally built and validated clinic CT-COVID AI tool. This effort directly supports the United Nations Sustainable Development Goals' number 3, Good Health and Well-Being, and allows sharing and transferring of knowledge to fight this devastating disease around the world.

Reduced Field-of-View Diffusion-Weighted Imaging in Histological Characterization of Rectal Cancer: Impact of Different Region-of-Interest Positioning Protocols on Apparent Diffusion Coefficient Measurements.

PURPOSE: To compare four region-of-interest (ROI) protocols for apparent diffusion coefficient (ADC) quantifications derived from reduced field-of-view diffusion-weighted imaging (rDWI) in histological characterization of rectal cancer. MATERIALS AND METHODS: Forty-nine patients with rectal cancer underwent rDWI at 3.0 T. Two readers independently performed mean and minimum ADC measurements using four different ROI positioning protocols (whole tumor volume [WTV], single-slice [SS], three-slices observer-based sampling [TSOB] and three-slices predefined sampling [TSPD]). Inter-observer variation was evaluated. Mean and minimum ADC values obtained from each method were compared in terms of different histological factors of rectal cancer, and their diagnostic abilities were assessed by receiver operating characteristic curve (ROC) analysis. The corresponding times for ADC measurements were recorded and compared between ROI methods. RESULTS: The inter-observer agreement was excellent for ADC values obtained by two readers using the four ROI methods (ICC range, 0.906-0.994). Mean and minimum ADC values by WTV method were significantly higher and lower than other methods respectively for both readers (P < 0.05). The AUCs of mean ADC measurements for assessment of well-differentiated tumors, T2 stage tumors and N0 status (no lymph node metastasis) (0.936, 0.840 and 0.714) were greater by WTV method than those by SS method (0.782, 0.761 and 0.677). The WTV method required longer measurement time than other ROI methods (P < 0.001). CONCLUSION: ADC measurements based on rDWI were influenced by ROI positioning protocols. rDWI technique had diagnostic value for histological characterization of rectal cancer, using WTV method with overall best inter-observer reproducibility, but with the longest measurement time.

Role of Chemical Exchange Saturation Transfer and Magnetization Transfer MRI in Detecting Metabolic and Structural Changes of Renal Fibrosis in an Animal Model at 3T.

OBJECTIVE: To investigate the value of combined chemical exchange saturation transfer (CEST) and conventional magnetization transfer imaging (MT) in detecting metabolic and structural changes of renal fibrosis in rats with unilateral ureteral obstruction (UUO) at 3T MRI. MATERIALS AND METHODS: Thirty-five Sprague-Dawley rats underwent UUO surgery (n = 25) or sham surgery (n = 10). The obstructed and contralateral kidneys were evaluated on days 1, 3, 5, and 7 after surgery. After CEST and MT examinations, 18F-labeled fluoro-2-deoxyglucose positron emission tomography was performed to quantify glucose metabolism. Fibrosis was measured by histology and western blots. Correlations were compared between asymmetrical magnetization transfer ratio at 1.2 ppm (MTRasym(1.2ppm)) derived from CEST and maximum standard uptake value (SUVmax) and between magnetization transfer ratio (MTR) derived from MT and alpha-smooth muscle actin (α-SMA). RESULTS: On days 3 and 7, MTRasym(1.2ppm) and MTR of UUO renal cortex and medulla were significantly different from those of contralateral kidneys (p < 0.05). On day 7, MTRasym(1.2ppm) and MTR of UUO renal cortex and medulla were significantly different from those of sham-operated kidneys (p < 0.05). The MTRasym(1.2ppm) of UUO renal medulla was fairly negatively correlated with SUVmax (r = -0.350, p = 0.021), whereas MTR of UUO renal medulla was strongly negatively correlated with α-SMA (r = -0.744, p < 0.001). CONCLUSION: CEST and MT could provide metabolic and structural information for comprehensive assessment of renal fibrosis in UUO rats in 3T MRI and may aid in clinical monitoring of renal fibrosis in patients with chronic kidney disease.

Assessment of renal fibrosis in a rat model of unilateral ureteral obstruction with diffusion kurtosis imaging: Comparison with α-SMA expression and 18F-FDG PET.

PURPOSE: To investigate the utility of diffusion kurtosis imaging (DKI) MRI for evaluation of renal fibrosis in rats with unilateral ureteral obstruction (UUO). METHODS: Twenty-five rats had UUO, and ten rats were subjected to sham operation as control. DKI was performed on a 3.0 T MRI scanner on days 1, 3, 5, and 7 after ligation. All rats then underwent 18F-FDG dynamic PET to evaluate unilateral renal function, followed by histological analysis to examine α-smooth muscle actin (α-SMA) expression. DKI metrics were assessed among the time points and between two sides, and compared with maximum standardized uptake value (SUVmax), serum levels of creatinine and urea, and fibrosis marker α-SMA. RESULTS: Mean kurtosis (MK) on day 7, axial kurtosis (Ka) on days 3 and 7, mean diffusivity (MD) on days 1, 3, 5, and 7, and fractional anisotropy (FA) on days 3, 5, and 7 of cortex and medulla between the UUO and contralateral sides were significantly different (all p < 0.05). Over the course of UUO progression, there were significant changes in Ka, MD and FA of medulla (all p < 0.05). FA of medulla was positively correlated with SUVmax (r = 0.641, p < 0.001), and MD of cortex was negatively correlated with urea (r = -0.534, p = 0.001). MD of cortex was negatively correlated with α-SMA on UUO sides (r = -0.710, p < 0.001). CONCLUSIONS: DKI shows the potential for noninvasive assessment of renal fibrosis and unilateral renal function induced by UUO.

MitoTEMPO Prevents Oxalate Induced Injury in NRK-52E Cells via Inhibiting Mitochondrial Dysfunction and Modulating Oxidative Stress.

As one of the major risks for urolithiasis, hyperoxaluria can be caused by genetic defect or dietary intake. And high oxalate induced renal epithelial cells injury is related to oxidative stress and mitochondrial dysfunction. Here, we investigated whether MitoTEMPO, a mitochondria-targeted antioxidant, could protect against oxalate mediated injury in NRK-52E cells via inhibiting mitochondrial dysfunction and modulating oxidative stress. MitoSOX Red was used to determine mitochondrial ROS (mtROS) production. Mitochondrial membrane potential (Δψm) and quantification of ATP synthesis were measured to evaluate mitochondrial function. The protein expression of Nox4, Nox2, and p22 was also detected to explore the effect of oxalate and MitoTEMPO on NADPH oxidase. Our results revealed that pretreatment with MitoTEMPO significantly inhibited oxalate induced lactate dehydrogenase (LDH) and malondialdehyde (MDA) release and decreased oxalate induced mtROS generation. Further, MitoTEMPO pretreatment restored disruption of Δψm and decreased ATP synthesis mediated by oxalate. In addition, MitoTEMPO altered the protein expression of Nox4 and p22 and decreased the protein expression of IL-6 and osteopontin (OPN) induced by oxalate. We concluded that MitoTEMPO may be a new candidate to protect against oxalate induced kidney injury as well as urolithiasis.