A transfer learning nomogram for predicting prostate cancer and benign conditions on MRI.

BACKGROUND: Deep learning has been used to detect or characterize prostate cancer (PCa) on medical images. The present study was designed to develop an integrated transfer learning nomogram (TLN) for the prediction of PCa and benign conditions (BCs) on magnetic resonance imaging (MRI). METHODS: In this retrospective study, a total of 709 patients with pathologically confirmed PCa and BCs from two institutions were included and divided into training (n = 309), internal validation (n = 200), and external validation (n = 200) cohorts. A transfer learning signature (TLS) that was pretrained with the whole slide images of PCa and fine-tuned on prebiopsy MRI images was constructed. A TLN that integrated the TLS, the Prostate Imaging-Reporting and Data System (PI-RADS) score, and the clinical factor was developed by multivariate logistic regression. The performance of the TLS, clinical model (CM), and TLN were evaluated in the validation cohorts using the receiver operating characteristic (ROC) curve, the Delong test, the integrated discrimination improvement (IDI), and decision curve analysis. RESULTS: TLS, PI-RADS score, and age were selected for TLN construction. The TLN yielded areas under the curve of 0.9757 (95% CI, 0.9613-0.9902), 0.9255 (95% CI, 0.8873-0.9638), and 0.8766 (95% CI, 0.8267-0.9264) in the training, internal validation, and external validation cohorts, respectively, for the discrimination of PCa and BCs. The TLN outperformed the TLS and the CM in both the internal and external validation cohorts. The decision curve showed that the TLN added more net benefit than the CM. CONCLUSIONS: The proposed TLN has the potential to be used as a noninvasive tool for PCa and BCs differentiation.

Multidetector computed tomography diagnosis of small bowel intussusceptions in adults: observations in a 10-year single-center study.

BACKGROUND: The etiologies of small bowel intussusception (SBI) in adults are varied. PURPOSE: To investigate multidetector computed tomography (MDCT) characteristics in adults with neoplastic and non-neoplastic SBI. MATERIAL AND METHODS: Clinical data and MDCT images diagnosed with SBI in adults from January 2010 to May 2020 were retrospectively reviewed. RESULTS: The study included a total of 71 patients. Forty-two patients had a combined total of 55 neoplastic intussusceptions, including 29 patients with benign tumors and 13 patients with malignant tumors. Twenty-nine patients had a combined total of 36 non-neoplastic intussusceptions, of which the condition was idiopathic in 23 patients and cased by non-neoplastic benign lesions in six patients. There were no significant differences in patient age or sex ratio in the neoplastic and non-neoplastic groups. In the non-neoplastic group the intussusceptions were shorter in length (3.6 cm vs. 13.2 cm, P<0.05) and smaller in transverse diameter (2.8 cm vs. 4.2 cm, P<0.05), and less likely to be associated with intestinal obstruction (2 vs. 18, P<0.05). The percentage of patients with multiple intussusceptions was greater in the neoplastic group (10/42, 23.8% vs. 4/29, 13.8%). In the non-neoplastic group only one lead point was detected (in a patient with Meckel's diverticulum), whereas lead points were detected in all 55 intussusceptions in the neoplastic group. CONCLUSION: There are differences in the clinical and MDCT manifestations of adult neoplastic and non-neoplastic SBIs. Whether a lead point is present or not has implications with regard to deciding on the most appropriate treatment and avoiding unnecessary surgery.

Methylation of Hypothalamic Tsc1-mTOR Signaling in Regulation of Obesity and Obesity Resistance.

The Tsc1-mTOR signaling pathway is often related to obesity, and epigenetic modification may lead to expression changes of obesity-related gene. Therefore, we aim to investigate the methylation of the Tsc1-mTOR signaling pathway in regulation of obesity susceptibility. Wistar rats were fed a normal diet or a high-fat diet to develop animal models. Protein and mRNA expression levels of Tsc1-mTOR signaling in the hypothalamus were determined by Western blot and quantitative real-time PCR. Methylation of Tsc1 gene promoter was detected by bisulfite genomic sequence. Both mRNA and protein expression levels of Tsc1 in DIO group hypothalamus were lower; mTOR and its downstream targets S6K1, 4EBP1, and S6 protein expression levels were higher than those of the DIO-R group and the chow group. The Tsc1 gene promoter methylation rate in the hypothalamus was 92.05 ± 3.07% in the DIO group, 87.27 ± 1.91% in the DIO-R group, and 88.18% ± 3.20% in the chow group, respectively, with significantly higher levels in the DIO group. Both the expression levels of Tsc1 gene promoter methylation and Tsc1-mTOR signaling pathway in the hypothalamus of DIO rats and DIO-R rats are different. These findings may shed light on the potential mechanism for the differentiation of obesity susceptibility.

CT and MRI imaging and interpretation of hepatic arterioportal shunts.

Hepatic arterioportal shunts (HAPS) occur due to organic or functional fistulization of blood flow between arterial hepatic vasculature and venous portal systems. It is a type of hemodynamic abnormality of the liver being observed increasingly with the use of temporal imaging modalities. HAPS occur due to other underlying hepatic abnormalities including the presence of an underlying tumor or malignancy. When a HAPS is present, the appearance of these abnormalities on imaging studies suggests an underlying abnormality, must be considered atypical even if asymptomatic, and warrants careful evaluation. Over time, and as a function of degree of fistulae, symptoms and potential life-threatening complications may arise from the HAPS. These systemic complications may include the development of portal hypertension, splenomegaly, as well as accelerated metastasis in patients with malignant tumors. This manuscript reviews common underlying conditions associated with HAPS and their radiologic interpretation.

Distribution and correlation of pancreatic gland size and duct diameters on MRCP in patients without evidence of pancreatic disease.

PURPOSE: To use MRCP to investigate age-related changes and gender differences of the pancreas and to correlate pancreatic gland size and duct diameter. METHODS: In this institutional review, board-approved, HIPAA-compliant study, 280 patients (age 20-88 years) without a history of pancreatic or liver disease who had undergone MRI/MRCP from 2004 to 2015 were identified. The anteroposterior size and main duct diameter of the pancreatic head, body, and tail were measured. The pancreatic gland and duct sizes were compared between genders, and among seven age subgroups (20-29, 30-39, 40-49, 50-59, 60-69, 70-79, 80-89). RESULTS: The pancreatic head and body were significantly larger in males than females (head, p < 0.01; body, p = 0.03), while the tail and the duct diameters of the pancreatic head, body, and tail showed no gender difference. As the age of male participants increased, there was an associated increase in size of the pancreatic gland initially (largest at age 50-59 (body) and 60-69 (head)), followed by subsequent decline in size thereafter. Additionally, the pancreatic duct diameter was found to increase gradually. In females, the size of the pancreatic gland decreased, while the diameter of the pancreatic duct increased with age. Moderate positive correlation for gland size and strong positive correlation for duct diameter among different pancreatic regions were found. Weak negative correlation was found between gland size and duct diameter. CONCLUSIONS: There are gender differences in the gland size of the pancreatic head and body. The pancreatic gland size increases until the sixth decade in males, with a more continuous decrease in gland size with age in females. Both males and females demonstrate a marked decrease in gland size after the eighth decade. The duct diameter increases with age in both males and females.

Electroacupuncture Reduces Weight in Diet-Induced Obese Rats via Hypothalamic Tsc1 Promoter Demethylation and Inhibition of the Activity of mTORC1 Signaling Pathway.

SUBJECT: The study aimed to investigate the mechanism of electroacupuncture reducing weight via tuberous sclerosis complex 1 (Tsc1) promoter methylation, inhibiting the mammalian target of rapamycin complex 1 (mTORC1) pathway. MATERIALS AND METHODS: Male Sprague-Dawley rats were divided into chow-fed group (chow group) or high-fat diet group (HF group) for 14 weeks. The obesity rats in HF group were randomly divided into electroacupuncture group (EA group) and diet-induced obesity (DIO) group, which received EA stimulation on bilateral ST25, RN12, SP6, and ST36 for 4 weeks or no further treatment, respectively. Methylation of the Tsc1 gene promoter and expression of agouti-related protein (AgRP), neuropeptide Y (NPY), and proopiomelanocortin (PoMC) were detected at the 18th week. RESULTS: At week 18, weight, body fat, and the body fat rate in DIO group were significantly higher than those of the chow and EA group. Compared with the chow group, the DIO group had increased methylation of the Tsc1 gene promoter and expression of mTORC1, AgRP, and NPY gene and decreased PoMC in the hypothalamus; after EA, methylation of the Tsc1 gene promoter, mRNA, and protein of the mTORC1 and expression of AgRP and NPY gene decreased and PoMC increased significantly. CONCLUSIONS: Our study could shed light on the potential pathway where EA exerts effects on the mechanism of EA treatment for obesity through the hypothalamic Tsc1 promoter demethylation and inhibition of the activity of mTORC1 signaling pathway.

Development and validation of a predictor of insufficient enhancement during the hepatobiliary phase of Gd-EOB-DTPA-enhanced magnetic resonance imaging.

Background Insufficient enhancement of liver parenchyma negatively affects diagnostic accuracy of Gd-EOB-DTPA-enhanced magnetic resonance imaging (MRI). Currently, there is no reliable method for predicting insufficient enhancement during the hepatobiliary phase (HBP) in Gd-EOB-DTPA-enhanced MRI. Purpose To develop a predictor for insufficient enhancement of liver parenchyma during HBP in Gd-EOB-DTPA-enhanced MRI. Material and Methods In order to formulate a HBP enhancement test (HBP-ET), clinical factors associated with relative enhancement ratio (RER) of liver parenchyma were retrospectively determined from the datasets of 156 patients (Development group) who underwent Gd-EOB-DTPA-enhanced MRI between November 2012 and May 2015. The independent clinical factors were identified by Pearson's correlation and multiple stepwise regression analysis; the performance of HBP-ET was compared to Child-Pugh score (CPS), Model for End-stage Liver Disease score (MELD), and total bilirubin (TBIL) using receiver operating characteristic (ROC) curve analysis. The datasets of 52 patients (Validation group), which were examined between June 2015 and Oct 2015, were applied to validate the HBP-ET. Results Six biochemical parameters independently influenced RER and were used to develop HBP-ET. The mean HBP-ET score of patients with insufficient enhancement was significantly higher than that of patients with sufficient enhancement ( P < 0.001) in both the Development and Validation groups. HBP-ET (area under the curve [AUC] = 0.895) had better performance in predicting insufficient enhancement than CPS (AUC = 0.707), MELD (AUC = 0.798), and TBIL (AUC = 0.729). Conclusion The HBP-ET is more accurate than routine indicators in predicting insufficient enhancement during HBP, which is valuable to aid clinical decisions.

Diffusion-weighted multiparametric MRI for monitoring longitudinal changes of parameters in rabbit VX2 liver tumors.

PURPOSE: To investigate the value of different quantitative models of diffusion-weighted multiparametric imaging (DW-MPI) including traditional as well as several advanced models for monitoring the longitudinal parameter changes in rabbit liver VX2 carcinoma and for correlating the perfusion-related imaging parameters to vascularity of tumor tissue. MATERIALS AND METHODS: Rabbit liver VX2 carcinoma was imaged by DW-MPI at the 2nd, 3rd, and 4th weeks after tumor implantation at 1.5T using two sets of b values ranging from 0 to 2000 s/mm(2) . Serial parameter changes of each model at three timepoints were compared. Univariate and multivariate regression analyses were carried out to analyze the ability of perfusion-related parameters, including apparent diffusion coefficient (ADC), perfusion fraction (f), and pseudodiffusion coefficient (D*), to predict mean microvessel density (MVD) as determined by quantitative histopathology. RESULTS: For the period from Week 2 to Week 4, the measurements of ADC, f, and KDKI illustrated a statistical difference (P = 0.000, P = 0.000, and P = 0.002, respectively), whereas the comparison of D, D*, DDKI , DSEM , and αSEM demonstrated no statistical significance. ADC and f showed highest correlation with MVD at Week 4 (r(2) = 0.307, P = 0.017, and r(2) = 0.402, P < 0.01, respectively). Multivariate analyses confirmed highest correlation of f and ADC with MVD at Week 4 (P = 0.001 and P = 0.002). CONCLUSION: ADC, f, and KDKI were identified as the most promising parameters for monitoring changes in rabbit liver VX2 carcinoma and f and ADC showed highest correlation with MVD. J. Magn. Reson. Imaging 2016;44:707-714.

Image quality evaluation of iterative model reconstruction on low tube voltage (80 kVp) coronary CT angiography in an animal study.

BACKGROUND: Iterative model reconstruction (IMR) is a newer generation iterative reconstruction method, but its value on coronary computed tomography (CT) angiography requires investigation. PURPOSE: To evaluate coronary image quality using IMR at a low-tube voltage of 80 kVp on coronary CT angiography in miniature pigs. MATERIAL AND METHODS: Five healthy miniature pigs underwent prospective electrocardiogram-gated coronary 256-slice CT angiography at 120 kVp and 80 kVp tube voltages, respectively. Filtered back projection (FBP) was used to reconstruct the 120 kVp standard-dose (SD) image sets (SD-FBP group), while iDose(4) and IMR were used to reconstruct the 80 kVp low-dose (LD) image sets (LD-iDose(4) and LD-IMR groups). Objective and subjective image qualities were compared among three groups. RESULTS: There were no significant differences in mean CT values of the ascending aorta, left main coronary artery, and right coronary artery between the LD-IMR and LD-iDose(4) groups (P > 0.05), but the values were significantly lower in the SD-FBP group than in those two groups (P < 0.05). The image noise in the LD-IMR group (21.5 ± 3.9 HU) was significantly lower than in the LD-iDose(4) (63.7 ± 9.8 HU) and SD-FBP (50.6 ± 4.6 HU) groups (P < 0.05). The signal-to-noise and contrast-to-noise ratios were significantly higher in the LD-IMR group than in the LD-iDose(4) and SD-FBP groups, respectively (P < 0.05). Subjective scoring of image noise, streak artifacts, and overall image quality indicated that the LD-IMR group was consistently superior to the LD-iDose(4) and SD-FBP groups (P < 0.05). CONCLUSION: IMR can significantly improve image quality at a low-tube voltage (80 kVp) with a 66% radiation dose reduction on coronary 256-slice CT angiography in miniature pigs.

Intravoxel incoherent motion diffusion-weighted MRI for characterizing regional variability and monitoring serial changes of parameters in rabbit VX2 liver tumors.

PURPOSE: To characterize intravoxel incoherent motion (IVIM) parameters in different regions of interest (ROIs) of rabbit VX2 liver tumors during a follow-up of 4 weeks. MATERIALS AND METHODS: This study was approved by the Institutional Animal Care and Use Committee. In 18 VX2 liver tumor-bearing rabbits, IVIM DW imaging was serially performed at the 2nd, 3rd, and 4th week separately after tumor implantation at 1.5T by using 12 b values. Three ROIs were drawn on the apparent diffusion coefficient (ADC) maps, including the whole lesion, the visibly most restricted diffusion area (MRDA), and peripheral area of tumor. For each ROI of tumors, serial changes in ADC and D, D*, and f at three timepoints were calculated. The differences of IVIM parameters in different ROIs of tumors at each timepoint were compared. RESULTS: Serial measurements of f in all three ROIs (P = 0.000, P = 0.002, and P = 0.000), ADC for the whole lesion, MRDA (P = 0.000 and P = 0.024), and D for MRDA (P = 0.001) from the 2nd to 4th week illustrated a statistical difference. The overall comparison of ADC (P = 0.000, P = 0.014, and P = 0.000), D (P = 0.000, P = 0.001, and P = 0.000), and f (P = 0.001, P = 0.001, and P = 0.000) for three tumor regions at weeks 2, 3, and 4 showed significant differences. D* in all three ROIs at each timepoint showed no significant differences (P = 0.612, P = 0.723, and P = 0.699; 95% confidence interval: 0.184-0.950). CONCLUSION: The peripheral area of tumor could be used as an alternative to the whole lesion as ROIs to assess the VX2 liver tumor with IVIM DW imaging, and in the period between weeks 2 and 3.

[Mechanism of Electropuncture for Reducing Diet-induced Obesity Rat Weight through Hypothala- mus TSC1 -mTOR Signal Pathway].

OBJECTIVE: To explore the mechanism of electropuncture (EA) for reducing diet-induced obesity (DIO) rat weight through tuberous sclerosis complex 1 (TSC1 )-mammalian target of rapa- mycin (mTOR) signal pathway in hypothalamus. METHODS: Forty male SD rats were randomly divided into the model group (n =30) and the normal control group (n =10). DIO rat model was prepared by high fat forage for 12 successive weeks. Successfully modeled 19 rats were further randomly divided into the model group (n =9) and the EA group (n =10). EA at Tianshu (ST25) , Sanyinjiao (SP6) , Zhongwan ( RN12) , Zusanli (ST36) was performed in the EA group, 5 successive days per week with a 2-day rest, 4 weeks in total. No intervention was given to rats in the model group and the normal control group. Body weight was observed in all rats. Methylation of TSC1 promotor was detected by bisulfite sequencing method. mRNA expression of mTOR in hypothalamus was detected by RT-PCR. RESULTS: After EA treatment body weight in the EA group were obviously reduced (P <0. 05). Compared with the normal control group, body weight was not statistically different between the model group and the EA group after treatment (P> 0. 05). Methylation rate of TSC1 promotor was higher in model group (94. 0% ±4. 5%) than in the normal control group (87. 0% ±3. 6%) and the EA group (87. 4% ±3. 9%) (P <0. 05). Expression of mTOR in the model group (1. 84 ±0. 51) was higher than that in the normal control group (1. 02 ±0. 22) and the EA group (1. 46 ±0. 29) (P <0. 05). CONCLUSION: EA could lower DIO rats' body weight by down-regulating methylation rate of TSC1 promotor and regulating expression of mTOR in hypothalamus.

[Tsc1 promoter methylation rate, mTOR expression in food-induced obese rat hypothalamus].

OBJECTIVE: To investigate the methylation rate of tuberous sclerosis complex 1 (Tsc1) promoter and expression of mammalian target of rapamycin (mTOR) in food-induced rat hypothalamus. METHODS: 16 male SD rats were divided into high fat diet induced group (8 rats) and normal control group (8 rats) feeding for 12 weeks. Body mass, mass of celiac fat, celiac fat/body mass were measured. Methylation of Tsc1 promoter, mRNA and protein expression of mTOR were detected by bisulfite sequencing method, RT-PCR and Western blot, respectively. RESULTS: Mass of celiac fat, celiac fat/body mass were higher in food-induced rat than that in control group. There were 11 methylation sites in SD rat hypothalamus. Obese group has significantly higher methylation rates (94.50% +/- 4.66%) than that of control group (86.60% +/- 3.49%) (P<0.002). The mRNA and protein expression of mTOR were noted lower in control group than in obese group (P<0.05). CONCLUSION: The increased methylation rate of Tsc1 promoter in food-induced rat hypothalamus and up-regulated expression of mTOR, downstream gene of Tsc1 may promote the obesity.

Successful treatment with continuous subcutaneous insulin infusion for allergy to human insulin and its analogs.

Allergy to insulin is rare. We report a type 2 diabetic patient who is allergic to human insulin and its analog preparations and successful treatment with continuous subcutaneous insulin infusion.

[Comparison on efficacy and safety of two regimens for treatment of type 2 diabetes mellitus: glargine plus metformin versus neutral protamine hagedorn plus metformin].

This is a work aimed to investigate the efficacy and safety of the combination of metformin with glargine or with neutral protamine Hagedorn in treatment of type 2 diabetes mellitus. Sixty such patients with poor glycemic control by oral antidiabetic drugs were included and divided into group A and group B. Thirty patients in group A were treated with glargine and metformin, and the other 30 patients in group B were treated with neutral protamine Hagedorn and metformin for 12 weeks. Fasting plasma glucose (FPG), postprandial glucose(PPG) and HbA1c were measured before and after the treatment. Hypoglycemia was also noted. At the end of the study, the levels of FPG, PPG and HbAlc were significantly lower than the baseline levels in the two groups (P < 0.05). At the 12th week, the percentage of HbAlc < 7% in group A was 53.3% and that in group B was 40.0%; statistically, there was no significant difference (P > 0.05). After the end of the treatment, there was no significant difference in that the percentage of HbA1c < 7% was 70.6% in group A and 62.5% in group B; the two groups' HbA1c levels were > or = 7%-9% at the baseline (P > 0.05). No sigificant difference in respect to the incidence rate of hypoglycemia in the two groups was noted (P > 0.05). In the cases of type 2 diabetes mellitus with poor glycaemic control by oral antidiabetic drug, glucose and HbA1c can be lowered further by the combination of metformin with glargine or with neutral protamine Hagedorn, the incidence rate of hypoglycemia is low. Metformin plus glargine or plus neutral protamine Hagedorn is a safe and effective therapeutic choice for type 2 diabetes mellitus cases with poor glycaemic control; moreover, metformin plus neutral protamine is a cheaper and effective choice.

[Effect of insulin plus rosiglitazone or metformin on serum N-terminal pro-brain natriuretic peptide in type 2 diabetes mellitus: a randomized-controlled study].

This study was aimed to compare the effect of insulin plus rosiglitazone with that of insulin plus metformin on the level of serum N-terminal pro-brain natriuretic peptide (NT-BNP) in patients with type 2 diabetes mellitus, and to find out whether serum NT-BNP can be used as an index for predicting heart failure induced by rosiglitazone in the cases of type 2 diabetes mellitus. Sixty type 2 diabetic patients were recruited and were randomly divided into two groups: group A (n = 30) received insulin plus rosiglitazone (4 mg/d) and group B (n = 30) received insulin plus metformin. The observations covered an 8-weeks' course of treatment. Serum NT-BNP was measured at the beginning and at the end of 8 weeks. The Before-After study revealed that the level of serum NT-BNP did not change apparently in the two groups (P >0.05). There was no remarkable difference in the level of serum NT-BNP between the two groups (P>0.05). There were 3 cases with edema in the group of insulin plus rosiglitazone, but none with heart failure; in these three cases, the mean serum NT-BNP level at the end of the treatment exhibited an increase of 108.99 fmol/ml when compared with that at the beginning. Neither insulin plus rosiglitazone nor insulin plus metformin had apparent effect on the level of serum NT-BNP in the patients with type 2 diabetes mellitus. The question of whether serum NT-BNP is a predictive index of heart failure awaits answers given by more observation on type 2 diabetes mellitus patients using rosiglitazone.

[Research on electricity frequency property of blood].

On the basis of our previous work, the electric frequency property of human blood in different components, in physiological state and in pathological state (diabetes) are tested and analyzed in the range of 1Hz-20MHz progressively. Among the different components of blood; the lowest electrical impedance is serum; the plasma and the whole blood gradually become larger, the blood corpuscle is the largest one. Otherwise, the negative phase of serum is the largest, the plasma and the whole blood are lower, and the blood corpuscle is the lowest. Here, the question is why the effect of the electric capacity of serum and plasma is the biggest in the condition of no cell and cell membrane; diabetes mellitus is an endocrine disorder in which blood changes obviously, the electric frequency property of the blood of diabetic patients changes markedly; the electrical impedance of blood decreases (more obviously with low frequency), the negative phase increases (more obviously with high frequency). These indicate that the increase of electric conductivity in diabetic patients' blood is due to electric capacitance conductivity that is related to the changes of cell membrane, deformation abilities and aggregation of RBC. Related experiments demonstrate again that with the progressing of research in the electric frequency property of blood, we may use the theory and method of electricity to examine some important characters of blood in a different way, and so to corroborate other tests and analyses.

[Measuring the electricity frequency properties of blood].

In order to understand the electricity frequency specialties of blood, we have developed a wide frequency electricity characteristic testing system and used it to test the amplitude frequency property and phase frequency property of the blood in different states and constituents at 1 Hz to 20 MHz. Further analysis on the results of tests helped us know some important properties of blood at even more microcosmic levels from a new angle. Meanwhile, some problems and considerations on the improvement of the electricity model of biotic tissue and blood were pointed out. (1) From 1 Hz to 5 KHz, the impedance of blood descended 99%. (2) Simple equivalent circuit of resistance and capacitance must be used in series equivalent but not in usual parallel connection equivalent. (3) Experiment indicated, equivalent circuits of blood need more analysis, because simple equivalent circuit of resistance and capacitance is liable to gross error. (4) When the three element model is used for measuring the resistance of inside liquid, capacitance of cell membrane and the resistance of outside liquid of blood, the three testing frequencies must be very similar.