Accuracy of 3.0T magnetic resonance angiography for the detection of arteriovenous fistula dysfunction in hemodialysis patients requiring interventional therapy: a prospective study.

Background: Color Doppler ultrasonography (CDUS) is feasible to detect arteriovenous fistula (AVF) dysfunction in hemodialysis patients but is not sufficient to map the structure of fistula required for interventions. This study is designed to evaluate the diagnostic accuracy of three-dimensional time-of-flight magnetic resonance angiography (TOF-MRA) at 3.0T versus CDUS for AVF dysfunction, by using digital subtraction angiography (DSA) as reference. Methods: This prospective study enrolled 68 consecutive patients with dysfunctional AVF who underwent both CDUS and TOF-MRA at Shanghai Sixth People's Hospital affiliated to Shanghai Jiao Tong University School of Medicine. The analysis of the dysfunctional AVFs was divided into three regions: the feeding artery, fistula and draining veins. In the whole- and per-regional-based analyses, two observers who were blinded to the clinical and DSA results independently analyzed all CDUS and TOF-MRA datasets. The image quality and stenosis severity of the lesions on TOF-MRA were evaluated. A receiver operating characteristic curve was applied to analyze the detection of AVF dysfunction with TOF-MRA. Results: A total of 204 vessel regions were evaluated. The whole-region-based image quality of TOF-MRA was poorer in patients with a total occlusion (1.8±0.8) than in those with stenosis (2.7±0.6, P<0.001). In the whole-region analyses, TOF-MRA had higher sensitivity [99.1% (94.6-100.0%) vs. 82.9% (74.6-89.0%), P<0.001] and similar specificity [93.1% (85.0-97.1%) vs. 94.3% (86.5-97.9%), P=0.755] than CDUS. The per-region-based analyses showed that TOF-MRA yielded higher sensitivity [fistula region, 98.1% (88.4-99.9%) vs. 80.8% (67.0-89.9%); P=0.004; draining vein region, 100.0% (92.5-100.0%) vs. 85.0% (72.9-2.5%); P=0.003] and similar specificity [fistula region, 88.2% (62.3-97.8%) vs. 88.2% (62.3-97.9%); P>0.99; draining vein region, 100.0% (59.8-100.0%) vs. 87.5% (46.7-99.3%); P>0.99] than CDUS. Sensitivity and specificity of TOF-MRA were comparable to those of CDUS in feeding artery region. Conclusions: TOF-MRA is a feasible and accurate method to display AVF dysfunction in hemodialysis patients, and this method might fulfill the endovascular treatment planning requirements.

Adversarial training for prostate cancer classification using magnetic resonance imaging.

Background: To use adversarial training to increase the generalizability and diagnostic accuracy of deep learning models for prostate cancer diagnosis. Methods: This multicenter study retrospectively included 396 prostate cancer patients who underwent magnetic resonance imaging (development set, 297 patients from Shanghai Jiao Tong University Affiliated Sixth People's Hospital and Eighth People's Hospital; test set, 99 patients from Renmin Hospital of Wuhan University). Two binary classification deep learning models for clinically significant prostate cancer classification [PM1, pretraining Visual Geometry Group network (VGGNet)-16-based model 1; PM2, pretraining residual network (ResNet)-50-based model 2] and two multiclass classification deep learning models for prostate cancer grading (PM3, pretraining VGGNet-16-based model 3; PM4: pretraining ResNet-50-based model 4) were built using apparent diffusion coefficient and T2-weighted images. These models were then retrained with adversarial examples starting from the initial random model parameters (AM1, adversarial training VGGNet-16 model 1; AM2, adversarial training ResNet-50 model 2; AM3, adversarial training VGGNet-16 model 3; AM4, adversarial training ResNet-50 model 4, respectively). To verify whether adversarial training can improve the diagnostic model's effectiveness, we compared the diagnostic performance of the deep learning methods before and after adversarial training. Receiver operating characteristic curve analysis was performed to evaluate significant prostate cancer classification models. Differences in areas under the curve (AUCs) were compared using Delong's tests. The quadratic weighted kappa score was used to verify the PCa grading models. Results: AM1 and AM2 had significantly higher AUCs than PM1 and PM2 in the internal validation dataset (0.84 vs. 0.89 and 0.83 vs. 0.87) and test dataset (0.73 vs. 0.86 and 0.72 vs. 0.82). AM3 and AM4 showed higher κ values than PM3 and PM4 in the internal validation dataset {0.266 [95% confidence interval (CI): 0.152-0.379] vs. 0.292 (95% CI: 0.178-0.405) and 0.254 (95% CI: 0.159-0.390) vs. 0.279 (95% CI: 0.163-0.396)} and test set [0.196 (95% CI: 0.029-0.362) vs. 0.268 (95% CI: 0.109-0.427) and 0.183 (95% CI: 0.015-0.351) vs. 0.228 (95% CI: 0.068-0.389)]. Conclusions: Using adversarial examples to train prostate cancer classification deep learning models can improve their generalizability and classification abilities.

Thin-slab maximum intensity projection of CT angiography for collateral score and clot burden score evaluation: comparison with conventional CT angiography.

BACKGROUND: This study aimed to exam the effects of thin-slab maximum intensity projection (TS-MIP) of computed tomography angiography (CTA) for collateral score (CS) and clot burden score (CBS) evaluation in patients with large-vessel-occlusion (LVO) stroke in the anterior circulation. METHODS: Of 241 consecutive patients with LVO stroke admitted to our center between August 2015 and June 2020, 187 patients were enrolled. CS and CBS were evaluated on conventional CTA and TS-MIP separately. Outcome at 90 days was classified as good if modified Rankin scale (mRS) was ≤2 and as poor if mRS was >2. The correlations between CS and CBS and clinical outcomes were assessed. Receiver operating characteristic (ROC) curve analysis was used to determine the diagnostic values of CS and CBS. Multivariate logistic regression analysis was performed to identify the independent predictors of 90-day good clinical outcomes. RESULTS: The correlation coefficient for clinical outcomes was significantly better for CS based on TS-MIP than that based on conventional CTA (-0.444 vs. -0.285, P=0.039); no significant difference was found in the CBS evaluation (TS-MIP: -0.356 vs. conventional CTA: -0.320, P=0.348). For predicting good clinical outcomes, TS-MIP-based CS was associated with larger area under the curve (AUC) (0.709 vs. 0.609, P=0.004) and higher sensitivity (69.1% vs. 42.0%, P=0.001) than CS based on CTA. In multivariable logistic regression analysis, the factors independently associated with good outcomes were National Institutes of Health Stroke Scale (NIHSS) score at admission (OR =1.147; P<0.001), TS-MIP-based CS (OR =0.326; P<0.001), final modified treatment in cerebral infarction (mTICI) score of 2b/3 (OR =0.098; P<0.001), and hemorrhagic transformation (OR =3.662; P<0.001). CONCLUSIONS: TS-MIP-CTA is superior to conventional CTA for evaluation CS and CBS, and TS-MIP-based CS may be a useful predictor of clinical outcome.

Calculation of Apparent Diffusion Coefficients in Prostate Cancer Using Deep Learning Algorithms: A Pilot Study.

BACKGROUND: Apparent diffusion coefficients (ADCs) obtained with diffusion-weighted imaging (DWI) are highly valuable for the detection and staging of prostate cancer and for assessing the response to treatment. However, DWI suffers from significant anatomic distortions and susceptibility artifacts, resulting in reduced accuracy and reproducibility of the ADC calculations. The current methods for improving the DWI quality are heavily dependent on software, hardware, and additional scan time. Therefore, their clinical application is limited. An accelerated ADC generation method that maintains calculation accuracy and repeatability without heavy dependence on magnetic resonance imaging scanners is of great clinical value. OBJECTIVES: We aimed to establish and evaluate a supervised learning framework for synthesizing ADC images using generative adversarial networks. METHODS: This prospective study included 200 patients with suspected prostate cancer (training set: 150 patients; test set #1: 50 patients) and 10 healthy volunteers (test set #2) who underwent both full field-of-view (FOV) diffusion-weighted imaging (f-DWI) and zoomed-FOV DWI (z-DWI) with b-values of 50, 1,000, and 1,500 s/mm2. ADC values based on f-DWI and z-DWI (f-ADC and z-ADC) were calculated. Herein we propose an ADC synthesis method based on generative adversarial networks that uses f-DWI with a single b-value to generate synthesized ADC (s-ADC) values using z-ADC as a reference. The image quality of the s-ADC sets was evaluated using the peak signal-to-noise ratio (PSNR), root mean squared error (RMSE), structural similarity (SSIM), and feature similarity (FSIM). The distortions of each ADC set were evaluated using the T2-weighted image reference. The calculation reproducibility of the different ADC sets was compared using the intraclass correlation coefficient. The tumor detection and classification abilities of each ADC set were evaluated using a receiver operating characteristic curve analysis and a Spearman correlation coefficient. RESULTS: The s-ADCb1000 had a significantly lower RMSE score and higher PSNR, SSIM, and FSIM scores than the s-ADCb50 and s-ADCb1500 (all P < 0.001). Both z-ADC and s-ADCb1000 had less distortion and better quantitative ADC value reproducibility for all the evaluated tissues, and they demonstrated better tumor detection and classification performance than f-ADC. CONCLUSION: The deep learning algorithm might be a feasible method for generating ADC maps, as an alternative to z-ADC maps, without depending on hardware systems and additional scan time requirements.

Advanced zoomed diffusion-weighted imaging vs. full-field-of-view diffusion-weighted imaging in prostate cancer detection: a radiomic features study.

OBJECTIVES: We aimed to compare the efficiency of prostate cancer (PCa) detection using a radiomics signature based on advanced zoomed diffusion-weighted imaging and conventional full-field-of-view DWI. METHODS: A total of 136 patients, including 73 patients with PCa and 63 without PCa, underwent multi-parametric magnetic resonance imaging (mp-MRI). Radiomic features were extracted from prostate lesion areas segmented on full-field-of-view DWI with b-value = 1500 s/mm2 (f-DWIb1500), advanced zoomed DWI images with b-value = 1500 s/mm2 (z-DWIb1500), calculated zoomed DWI with b-value = 2000 s/mm2 (z-calDWIb2000), and apparent diffusion coefficient (ADC) maps derived from both sequences (f-ADC and z-ADC). Single-imaging modality radiomics signature, mp-MRI radiomics signature, and a mixed model based on mp-MRI and clinically independent risk factors were built to predict PCa probability. The diagnostic efficacy and the potential net benefits of each model were evaluated. RESULTS: Both z-DWIb1500 and z-calDWIb2000 had significantly better predictive performance than f-DWIb1500 (z-DWIb1500 vs. f-DWIb1500: p = 0.048; z-calDWIb2000 vs. f-DWIb1500: p = 0.014). z-ADC had a slightly higher area under the curve (AUC) value compared with f-ADC value but was not significantly different (p = 0.127). For predicting the presence of PCa, the AUCs of clinical independent risk factors model, mp-MRI model, and mixed model were 0.81, 0.93, and 0.94 in training sets, and 0.74, 0.92, and 0.93 in validation sets, respectively. CONCLUSION: Radiomics signatures based on the z-DWI technology had better diagnostic accuracy for PCa than that based on the f-DWI technology. The mixed model was better at diagnosing PCa and guiding clinical interventions for patients with suspected PCa compared with mp-MRI signatures and clinically independent risk factors. KEY POINTS: • Advanced zoomed DWI technology can improve the diagnostic accuracy of radiomics signatures for PCa. • Radiomics signatures based on z-calDWIb2000 have the best diagnostic performance among individual imaging modalities. • Compared with the independent clinical risk factors and the mp-MRI model, the mixed model has the best diagnostic efficiency.

Atherosclerosis in intracranial or extracranial vessels in diabetic patients and the association with stroke subtype.

BACKGROUND: Diabetes mellitus (DM) is associated with increased prevalence and severity of atherosclerosis. This study aimed to assess the prevalence and location of atherosclerosis in intracranial and extracranial vessels in diabetic patients and to investigate their association with ischemic stroke subtype. METHODS: Diabetes patients (n=128) and nondiabetic patients (n=195) were enrolled. Brain MRI, MR angiography, and digital subtraction angiography (DSA) imaging findings in the two groups were retrospectively compared. The characteristics of atherosclerosis (prevalence, location, severity) and collateral flow in diabetic and nondiabetic patients and their association with stroke subtype were analyzed. RESULTS: Atherosclerosis in extracranial vessels was more common in diabetes patients than in nondiabetic patients (43.8% vs. 23.1%; P<0.001). Symptomatic stenoses were commonly in the proximal internal carotid artery (ICA) and proximal vertebral artery (pVA). Diabetes patients were more likely to have lacunar infarction (49.2% vs. 32.3%; P=0.002) and less likely to have large artery infarct (36.7% vs. 48.2%; P=0.042). DM (OR, 2.03; 95% CI, 1.96-4.30; P=0.006) and age >65 years (OR, 2.55; 95% CI, 1.24-5.22; P=0.011) were independent risk factors for lacunar infarct. Diabetes patients with symptomatic extracranial stenosis or occlusion, combined with good collateral circulation, had significantly higher risk of lacunar infarction than nondiabetic patients (47.8% vs. 30.5%; P=0.045). CONCLUSIONS: DM aggravates the severity of extracranial atherosclerosis. Lacunar stroke is relatively common in diabetic patients and could even be due to large artery disease (LAD).

CARM1 Methylates GAPDH to Regulate Glucose Metabolism and Is Suppressed in Liver Cancer.

Increased aerobic glycolysis is a hallmark of cancer metabolism. How cancer cells coordinate glucose metabolism with extracellular glucose levels remains largely unknown. Here, we report that coactivator-associated arginine methyltransferase 1 (CARM1 or PRMT4) signals glucose availability to glyceraldehyde-3-phosphate dehydrogenase (GAPDH) and suppresses glycolysis in liver cancer cells. CARM1 methylates GAPDH at arginine 234 (R234), inhibiting its catalytic activity. Glucose starvation leads to CARM1 upregulation, further inducing R234 hypermethylation and GAPDH inhibition. The re-expression of wild-type GAPDH, but not of its methylation-mimetic mutant, sustains glycolytic levels. CARM1 inhibition increases glycolytic flux and glycolysis. R234 methylation delays tumor cell proliferation in vitro and in vivo. Compared with normal tissues, R234 is hypomethylated in malignant clinical hepatocellular carcinoma samples. Notably, R234 methylation positively correlates with CARM1 expression in these liver cancer samples. Our findings thus reveal that CARM1-mediated GAPDH methylation is a key regulatory mechanism of glucose metabolism in liver cancer.

Integrated application of antegrade and retrograde recanalization for femoral-popliteal artery chronic total occlusions: outcomes compared with antegrade recanalization.

BACKGROUND: To investigate the efficacy of retrograde recanalization for chronic total occlusion (CTO) of femoral-popliteal artery in patients with peripheral arterial disease. METHODS: In this single-center retrospective study, all patients who had undergone endovascular recanalization for femoral-popliteal CTOs at our center from June 2011 to October 2014 were included. Patients' demographics, immediate and follow-up outcomes were analyzed. RESULTS: A total of 205 patients with 238 CTOs were enrolled. In total, successful recanalization was achieved in 228 CTOs (95.8%). The antegrade procedure was successful in 196 CTOs. The retrograde procedure was successfully performed in 32 CTOs after failed antegrade procedure. Ankle-brachial index increased from 0.48±0.18 to 0.79±0.16 in antegrade group vs. 0.41±0.13 to 0.76±0.13 in retrograde group (P=0.438). Pulse score increased from 0.48±0.50 to 2.30±0.76 in antegrade group vs. 0.48±0.51 to 2.30±0.79 in retrograde group (P=0.771). At 12 and 24 months, primary patency rate was 86.2% (169/196) and 51.5% (101/196) in the antegrade group, and 75.0% (24/32) and 43.8% (14/32) in the retrograde group, respectively (P=0.346). Kaplan-Meier analysis showed limb salvage rates of 85.7% in the antegrade group vs. 78.1% in the retrograde group (P=0.198). CONCLUSIONS: Retrograde recanalization is effective for CTO of femoral-popliteal artery after the failure of an antegrade procedure; immediate outcomes and mid-term patency and limb salvage rate are comparable with that of antegrade procedure.

Development of a web-based liver cancer prediction model for type II diabetes patients by using an artificial neural network.

BACKGROUND: Diabetes mellitus is associated with an increased risk of liver cancer, and these two diseases are among the most common and important causes of morbidity and mortality in Taiwan. PURPOSE: To use data mining techniques to develop a model for predicting the development of liver cancer within 6 years of diagnosis with type II diabetes. METHODS: Data were obtained from the National Health Insurance Research Database (NHIRD) of Taiwan, which covers approximately 22 million people. In this study, we selected patients who were newly diagnosed with type II diabetes during the 2000-2003 periods, with no prior cancer diagnosis. We then used encrypted personal ID to perform data linkage with the cancer registry database to identify whether these patients were diagnosed with liver cancer. Finally, we identified 2060 cases and assigned them to a case group (patients diagnosed with liver cancer after diabetes) and a control group (patients with diabetes but no liver cancer). The risk factors were identified from the literature review and physicians' suggestion, then, chi-square test was conducted on each independent variable (or potential risk factor) for a comparison between patients with liver cancer and those without, those found to be significant were selected as the factors. We subsequently performed data training and testing to construct artificial neural network (ANN) and logistic regression (LR) prediction models. The dataset was randomly divided into 2 groups: a training group and a test group. The training group consisted of 1442 cases (70% of the entire dataset), and the prediction model was developed on the basis of the training group. The remaining 30% (618 cases) were assigned to the test group for model validation. RESULTS: The following 10 variables were used to develop the ANN and LR models: sex, age, alcoholic cirrhosis, nonalcoholic cirrhosis, alcoholic hepatitis, viral hepatitis, other types of chronic hepatitis, alcoholic fatty liver disease, other types of fatty liver disease, and hyperlipidemia. The performance of the ANN was superior to that of LR, according to the sensitivity (0.757), specificity (0.755), and the area under the receiver operating characteristic curve (0.873). After developing the optimal prediction model, we base on this model to construct a web-based application system for liver cancer prediction, which can provide support to physicians during consults with diabetes patients. CONCLUSION: In the original dataset (n=2060), 33% of diabetes patients were diagnosed with liver cancer (n=515). After using 70% of the original data to training the model and other 30% for testing, the sensitivity and specificity of our model were 0.757 and 0.755, respectively; this means that 75.7% of diabetes patients can be predicted correctly to receive a future liver cancer diagnosis, and 75.5% can be predicted correctly to not be diagnosed with liver cancer. These results reveal that this model can be used as effective predictors of liver cancer for diabetes patients, after discussion with physicians; they also agreed that model can assist physicians to advise potential liver cancer patients and also helpful to decrease the future cost incurred upon cancer treatment.

Transdermal application of azithromycin-amlodipine-heparin gel enhances survival of infected random ischaemic flap.

BACKGROUND: Flap necrosis is generally regarded as the result of vasospasm, thrombosis, and infection. METHODS: To improve skin flap survival and lower the risk of side effects due to systemic drug delivery, we formulated and evaluated compound gels for transdermal application. The transdermal delivery of 1% azithromycin (AZM), 0.5% amlodipine besylate (AB), and 300 IU/g low molecular weight heparin (LMWH) in compound gels, singly or in combinations, was measured across rat skin in vitro. The effects of AB and LMWH on flap blood circulation was investigated using fluorescein angiography, by transdermally applying the gel onto the surface of an in vivo ischaemic flap rat model; concentrations of the drugs were detected in both blood plasma and flap tissue at assigned timepoints. Finally, infected ischaemic flaps were treated to evaluate their anti-inflammatory effects and sizes of flap survival area. RESULTS: Each drug efficiently penetrated the in vitro skin in a time-dependent manner. In the in vivo ischaemic flaps, AB or LMWH increased the blood supply. All gel formulations that included AZM were associated with less flap inflammation. The surviving areas after treatment with AZM+LMWH or AZM+AB were significantly larger than that treated with the AZM-only gel, and the largest surviving area was that treated with AZM+AB+LMWH. Gels containing no AZM could not decrease flap inflammation or increase flap survival. CONCLUSION: Transdermal application of a compound gel with AZM, AB, and LMWH combined is a promising method to prevent and treat flap infection, improve blood circulation, and increase the survival of infected ischaemic flaps.

Pretreatment of low-abundance peptides on detonation nanodiamond for direct analysis by matrix-assisted laser desorption/ionization time-of-flight mass spectrometry.

Detonation nanodiamond (dND) was firstly employed as adsorbent for pretreatment of peptides in dilute/contaminated sample solution. Detonation nanodiamond showed high efficiency for isolating and enriching peptides in a wide pH range. Remarkably, good tolerance capability toward salts and detergents could be achieved by using dNDs. Due to the inherent specificities of dNDs, dND-bound peptides could be directly analyzed by MALDI-TOF MS, so as to avoid the elution step and reduce sample loss. This pretreatment method also exhibited a better performance for protein identification compared to solvent evaporation and Ziptip pretreatment approach.

Proteome analysis of human pancreatic ductal adenocarcinoma tissue using two-dimensional gel electrophoresis and tandem mass spectrometry for identification of disease-related proteins.

A comparative proteomic approach has been used to identify and analyze proteins related to pancreatic cancer. Proteomes of eight pairs of clinical pancreatic ductal adenocarcinoma (PDAC) tissue samples and samples of normal adjacent tissue were obtained by two-dimensional gel electrophoresis (2DE). Comprehensive analysis of proteins was focused on total protein spots for which there were statistical differences between the two groups. Proteins were identified by peptide mass fingerprinting with tandem mass spectrometry (MS-MS). Western blotting and immunohistochemistry (IHC) were also performed to verify the expression of some candidate proteins. Thirty protein spots were identified, including proteases, antioxidant proteins, signal-transduction proteins, calcium-binding proteins, structural proteins, chaperones, and others. Western blotting and IHC confirmed up-regulated expression of two candidate proteins, nucleotide diphosphatase kinase (NDPK) and annexin II, in tumorous tissues. These results suggest that combination of 2DE with MS is an effective strategy for discovery of differently expressed proteins in PDAC which may be molecular markers for diagnosis or therapeutic targets.

A glycoproteome database of normal human liver tissue.

PURPOSE: To extensively investigate the glycoproteins of normal human liver tissue, constructing the glycoprotein profile and database of the normal human liver tissue. METHODS: The total proteins were extracted from the normal human liver tissue and then subjected to two-dimensional electrophoresis (2-DE). Finally, 2-DE gels were stained according to the methods of multiplexed proteomics (MP) technology. Glycoprotein spots were excised from 2-DE gel and then characterized by matrix assisted laser desorption/ionization-time of flight mass spectrometry (MALDI-TOF-MS). RESULTS: The PDQuest software detected 1,011 glycoprotein spots and 1,923 total protein spots in the 2-DE gels of sample from the normal human liver tissue. Furthermore, 116 species of glycoproteins were successfully identified via peptide mass profiling using MALDI-TOF-MS/MS and annotated to our databases. In addition, we also applied bioinformatics softwares to predict N- or O-glycosylation sites of identified glycoproteins. CONCLUSION: This study demonstrates the feasibility of a novel technological platform to contruct glycoprotein databases. These results lay the foundation for future physiological and pathological studies of the human liver.