Objective:To investigate the impact of the deep learning reconstruction algorithm TrueFidelity TM for Gemstone Spectral Imaging (TF-GSI) and the adaptive statistical iterative reconstruction algorithm (ASiR-V, hereinafter referred to as ASiR-V) based on phantom and animal models on the image quality of dual-energy CT images. Methods:GE Revolution Apex CT was used to scan the ACR 464 phantom and a mouse model of gastric cancer with lymph node metastasis ( n=16). TF-GSI and ASiR-V were separately used to reconstruct middle and high-grade images (TF-GSI-M, TF-GSI-H, ASiR-V-50%, and ASiR-V-100%) on the phantom and mouse based on virtual monoenergetic images at 70 keV. The task transfer function (TTF) of bone and acrylic, image noise power spectrum (NPS), and detectability index (d′) of the phantom images were evaluated. One-way ANOVA analysis was used to compare the image noise, signal-to-noise ratio (SNR) and contrast-to-noise ratio (CNR) for brain and liver on images of mice. The consistency of the two reconstruction-algorithm images (TF-GSI-H and ASiR-V100%) in the detection of small lesions by two radiologists (A and B) was evaluated using kappa test. Results:In terms of the phantom, the TF-GSI-H group had the best performance in TTF, NPS, and d′. Compared to ASiR-V-100%, the TTF50% of bone and acrylic in the TF-GSI-H group increased by 2.4% and 8.9%, respectively; the NPS peak decreased by 54.1%, compared to ASiR-V-100%; the d′ of bone and acrylic in the TF-GSI-H group relative to ASiR-V-100% increased by 52.7% and 59.5%, respectively. The TF-GSI group had reduced image noise compared to the ASiR-V group, and both SNR and CNR of the two tissues increased, but the differences between the groups were not statistically significant (all P>0.05). The two reconstruction-algorithm images showed good consistency in image evaluation by the two radiologists (A, Kappa=0.875, P<0.001; B, Kappa=0.625, P=0.012). In terms of the detection of micro-metastases in mice, the TF-GSI group outperformed the ASiR-V group (average accuracy: 83.5% vs 71.9%; average sensitivity: 77.8% vs 61.2%; average specificity: 85.7% vs 85.7%). Conclusion:Compared with iterative reconstruction algorithm, the DLIR algorithm showed improved spatial resolution, reduced image noise, and enabled detectability of micro-lesion for images from dual-energy CT.

目的:探讨多排CT (MDCT)在鉴别胰腺实性型浆液性囊腺瘤(sSCA)与富血供神经内分泌肿瘤(pNET)中的价值.方法:回顾性分析2014年1月至2020年2月经上海交通大学医学院附属瑞金医院手术病理证实为胰腺sSCA(13例)及pNET(26例)的临床、上腹部CT资料.采用x2检验比较2组患者定性参数(性别、病灶部位、形态等),采用两独立样本t检验比较2组患者定量参数(病灶大小、平扫CT值、动脉期CT值等),对其中差异有统计学意义的定量参数绘制ROC曲线并比较曲线下面积(AUC)以评价其诊断效能.结果:sSCA (12/13)相较于pNET (13/26)平扫多呈相对低密度改变,sSCA平扫、门脉期CT值[(29.4±3.7)HU、(107.3±13.1) HU)]低于pNET[(43.6±6.5) HU、(123.4±18.4) HU],而动脉期、门脉期强化率[(4.1±1.3)×100％、(2.8±0.8) ×100％]均高于pNET[(2.9±0.9)×100％、(1.9±0.5)×100％],差异均有统计学意义(P＜0.05).ROC曲线显示平扫CT值所对应AUC (0.959)高于增强动脉期强化率(0.781)、门脉期CT值(0.750)、门脉期强化率(0.831),差异有统计学意义(P＜0.05),平扫CT值≤36.8 HU诊断sSCA的灵敏度、特异度分别为100％、92.3％.结论:胰腺sSCA与富血供pNET的CT表现常易混淆,平扫CT值对两者鉴别诊断具有较大价值.

Objective:To explore the optimal keV for the visualization of rectal cancer and to investigate its diagnostic performance in the preoperative T staging of rectal cancer using virtual monoenergetic image (VMI) on a novel dual-layer spectral detector CT.Methods:Totally 43 patients with pathologically confirmed rectal cancer were involved in this retrospective study in Ruijin Hospital, Shanghai Jiao Tong University School of Medicine from May to October 2019. All the patients underwent novel dual-layer spectral detector CT scanning within 1 week before surgery. The 40, 50, 60, 70 keV VMI and 120 kVp conventional polyenergetic image (PI) were reconstructed based on the venous phase imagings. The signal-to-noise ratio (SNR) and contrast-to-noise ratio (CNR) of the rectal cancer were measured and the image quality was scored using a 5-point scale to obtain the best keV VMI. Taking pathological results as golden standard, the accuracy rate of T staging was calculated and compared between the best keV VMI and 120 kVp PI. The CT attenuation, SNR and CNR were compared with one-way ANOVA analysis among the 5 groups of images.Image quality scores were assessed with Friedman test,and the accuracy rate was compared by McNemar test or Fisher exact test.Results:The SNR and CNR of rectal cancer at 40-70 keV VMI was better than 120 kVp PI ( P<0.05), the SNR at 40 keV VMI was the highest among all the VMI ( P<0.05), the CNR at 40 keV VMI was better, but there was no significant difference between 40 and 50 keV VMI ( P>0.05). The image scores at 40-60 keV VMI were all significantly superior to those of the 120 kVp PI ( P<0.001). 40 keV VMI was the best image. The accuracy rate of 40 keV VMI (67.4%, 29/43) was higher than that of 120 kVp PI (48.8%, 21/43) in the T stage ( P=0.008). The 40 keV VMI had a higher diagnostic accuracy rate (83.7%, 36/43) and specificity （58.8%, 10/17） in diagnosis of locally advanced rectal cancer (T3-4, P=0.016, 0.031). Conclusion:The image quality of novel dual-layer spectral detector CT at 40 keV VMI is good, which can improve the accuracy rate of preoperative T staging of rectal cancer.

目的:探讨基于可变形图像配准(deformableregistration algorithms,DRA)算法的心肌形变分析在肥厚型心肌病(hypertrophic cardiomyopathy,HCM)中的诊断及预后预测价值.方法:回顾性收集本院确诊的HCM患者23例,另招募23名年龄性别匹配的健康志愿者(对照组),均行心脏MRI扫描.HCM组按有无心肌延迟强化(late gadoli-nium enhancement,LGE)分为LGE(+)组和LGE(-)组.采用基于DRA算法的TrufiStrain软件进行心肌形变分析,测量整体径向应变峰值(global peak radial strain,GPRS)、整体环向应变峰值(global peak circumferential strain,GPCS)和整体纵向应变峰值(global peak longitudinal strain,GPLS)以及心尖部、中部和基底部的径向应变峰值(peak radial strain,PRS)、环向应变峰值(peak circumferential strain,PCS)和纵向应变峰值(peak longitudinal strain,PLS),探讨心肌形变参数在HCM中的诊断效能及其与心室壁厚度、LGE间的关系.结果:除心尖部PCS外,HCM组的GPRS、GPCS、GPLS及基底部和中部PRS、PCS、PLS,以及心尖部的PRS、PLS均明显低于对照组(P<0.05),而HCM组的整体及各部分3个方向(环向、纵向、径向)的应变率比值均明显小于对照组(P<0.05),且LGE(-)组及对照组各部分的PRS、PCS、PLS及3个方向的舒张早期和晚期应变率比值[径向应变率比值(ratio of radial strain rate during early and late diastole,REL),环向应变率比值(ratio of circumferential strain rate during early and late diastole,CEL)和纵向应变率比值(ratio of longitudinal strain rate during early and late diastole,LEL)]均大于LGE(+)组(P<0.05).GPRS、GPCS、GPLS、REL、CEL、LEL均与左心室心肌最大厚度明显相关(P<0.05).GPRS、GPCS、GPLS、REL、CEL及LEL诊断HCM的曲线下面积分别为0.686、0.905、0.921、0.972、0.974及0.917,GPRS、GPCS、GPLS、REL、CEL及LEL预测HCM患者是否有LGE的曲线下面积均大于0.871.结论:基于DRA的心肌形变分析在诊断HCM、预测HCM患者发生LGE中具有重要价值,且与心肌厚度密切相关.

Purpose: To validate a software prototype based on pharmacokinetic and physiologic models, which aimed to optimize contrast media (CM) injection protocol for coronary computed tomography angiography (CCTA) according to patient's test-bolus data. Methods: All the patients underwent CCTA in our hospital during June-Spetember 2016 were randomly assigned into group A (with bolus tracking technique), group B (with test bolus technique) or group C (with prototype software). The desired enhanced attenuation value was set as 350 HU. Attenuation values were measured manually by a radiologist. All the data were analyzed with ANOVA and Levene test by SPSS (version 18.0) (SPSS Inc, Chicago, Ill). Results: A total of 199 cohort patients were recruited. The enhancement level of group A was the highest among these three groups, and was higher than the targeted attenuation (350 HU). Compared with group B, the group C obtained better enhancement effect in middle (S2) and distal segment (S3) of right coronary artery and similar enhancement effect in the other segments. Meanwhile, the amount of CM used in group C was the least. The standard variation of enhancement effect of group C was the lowest except S2 and S3, which meant group C had the least inter-individual enhancement variations except S2 and S3. Conclusions: Compared with traditional injection protocol, the prototype of software can be used to obtain better attenuation (closer to the targeted attenuation) with significantly fewer CM, while inter-individual enhancement variations of coronary artery segments are the least except the middle and distal segment of right coronary artery.

Objective: To evaluate the accuracy of the preoperative local-regional staging of the rectal cancer by a two-dimensional (2D) T2-weighted TSE sequence and a three-dimensional (3D VISTA) T2-weighted turbo spin-echo (TSE) magnetic resonance (MR) sequence. Materials and Methods: This retrospective study was conducted of 33 patients who diagnosed rectal cancer pathologically from January to April in 2016. Two radiologists independently assessed the radiologic images for T and N staging. All patients were performed with both 2D-T2WI MRI and 3D VISTA MRI of the rectum by Philips 3.0T MR imager. The accuracy and consistence with pathological results of TN staging using 2D and 3D sequences were calculated using SPSS 23. Results: in 33 cases, T staging accuracy values of 2D and 3D VISTA data were 70% and 85% , respectively (Kappa value were 0.576 and 0.790) for reviewer 1 (P = 0.274) and 73% and 82% (Kappa value were 0.618 and 0.746) for reviewer 2 (P = 0.454), The agreement of VISTA T sequences between 3D staging and pathologic results was excellent and better than that of 2D sequences; N category lesion staging accuracy values for 2D and 3D VISTA data were 58%and 70%, respectively (Kappa value were 0.252 and 0.407) for reviewer 1 (P = 0.182) and 58% and 67% (Kappa value were 0.267 and 0.376) for reviewer 2 (P = 0.218). The agreement of N staging 2D and VISTA 3D staging with pathological results were both poor agreement. Conclusion: Preoperative staging with conventional 2D MR imaging sequence and 3D VISTA T2-weighted imaging protocols for rectal cancer patients showed no significant differences in accuracy of TN staging. However, compared to 2D sequence, 3D VISTA sequence has its advantages in preoperative MR staging.

To assess the value of diffusion-weighted imaging sequence in the preoperative staging of rectal cancer.Between January and December 2007,sixty-seven patients with operable rectal cancer underwent preoperative MR imaging with single-shot echo plannar imaging(EPI) diffusion-weighted sequences(b=600,1000s/mm2).All of them were compared with pathologic staging.The apparent diffusion coefficients(ADCs) of the tumors、lymph nodes and lip in each patient were measured to evaluate the correlation between ADC and pathological classification of each tumor,and between benign and malignant lymph nodes.The mean ADC values of tumor and lymph nodes were decreases following the ascending of b values.No overt correlation can be pointed out between ADC and pathological classification of each tumor,and there were correlation between benign and malignant lymph nodes of the mean ADC values.Diffusion-weighted imaging is a rapid and feasible method in detecting rectal cancer and lymph nodes,the mean ADC values give help for distinguishing between benign and malignant lymph nodes,DWI may be regarded as a conventional sequence in the preoperative staging of rectal cancer.

随着生活水平的提高,直肠癌在我国的发病率明显提高[1],随之而来的是对诊疗水平要求的提升.据国内外多项研究表明,直肠癌患者是否伴淋巴结转移,是否进行术前辅助放化疗、手术方式及范围的选择,是否进行术后放疗,局部复发率以及远端转移率等密切相关,是一项重要的影响预后的因素[2].而常规的成像方式对淋巴结状态的评估差强人意[3].目前磁共振成像在直肠癌的淋巴结状态判断方面日益受到重视.本文就其在直肠癌淋巴结转移中应用及展望进行综述.

Background and purpose:The mutation of K-rasG12D bring about pancreatic intraepithelial neoplasia(PanIN),which is thought to be a precursor lesion of pancreatic ductal adenocarcinoma via the research in gene targeted mice model.PanIN with the silencing of Smad4 could progress to invasive adenocarcinomas.The change about the genes expression profile between PanIN cells and PanIN cells with Smad4 silencing(PanIN-S cells) which promoted PanIN cells malignant transformation still remains poorly understood.In our previous research,PanIN-S cells were established by stable transfection with lentiviral mediated Smad4 RNA interference.In the present study,we aimed to explore the gene differentially expressed genes between PanIN cells and PanIN-S cells.Methods:A gene microarray(Agilent) with 41 174 transcripts?was employed to examine the gene differential expression in PanIN cells and PanIN-S cells.The differentially expressed genes were identified and subjected to Real-time PCR analysis.Results:A large number of differentially expressed genes whose fold change was larger than 2 were examined,embracing Fut9,CXCR4,MMP2,PIK3C2G,Cplx4,Lemd1,TIMP4,Reln,PITX2 and so on.The fold change of Real-time PCR of CXCR4,MMP2,PIK3C2G,TIMP4,PITX2 was consistent with that of gene expression profiling.Conclusion:The outcome of the gene expression profiling indicated many differentially expressed genes associated with PanIN cells and PanIN-S cells.CXCR4,PIK3C2G may be related to the pancreatic tumorigeness via angiogenesis.MMP2,TIMP4 may be bound up with adhesion,invasion and migration of cancer cells.The expression levels of PITX2 may be associated with prognosis of pancreatic cancer.

Pancreatic cancer is a lethal disease with low early detection rate and poor prognosis.Pancreatic intraepithelial neoplasia(PanIN) is the most common non-invasive precursor lesion of pancreatic ductal adenocarcinoma(PDAC).The identification of PanIN provides opportunities for early detection and potential therapeutic intervention before the development of invasive PDAC.In this article,we reviewed the clinical and pathological characteristics,as well as the genetic changes of PanIN.

Background: Inactivation of tumor suppressor gene Smad4 can promote the malignant transformation of pancreatic intraepithelial neoplasia(PanIN) cells initiated by Kras mutation,but its mechanism has not yet been clarified.Aims: To investigate the alteration of gene expression profile of Smad4-silenced PanIN cells(PanIN-S cells),and to assess the possible mechanism of Smad4 gene silencing induced proliferation and malignant transformation of PanIN cells.Methods: PanIN-S cells were obtained by silencing the endogenous Smad4 gene in PanIN cells through RNA interference.Differential gene expression profile of PanIN cells and PanIN-S cells were detected by mouse pan-gene chip array analysis,and real-time fluorescent quantitative RT-PCR was performed to further identify the differentially expressed cell cycle-related genes detected by gene chip array amalysis.Results: Gene chip array analysis detected 237 differentially expressed genes,of which 148 were up-regulated and 89 were down-regulated in PanIN-S cells,mostly involved in cell cycle,proliferation,apoptosis,adhesion,transcription,etc.Real-time fluorescent quantitative RT-PCR demonstrated the identity of differential expression of cell cycle-related genes p27,p19 and cyclin D1 detected by gene chip array analysis in PanIN cells and PanIN-S cells.Conclusions: Gene expressions of p27,p19 and cyclin D1 in PanIN-S cells alter significantly,which may be involved in the proliferation and malignant transformation of PanIN-S cells.

OBJECTIVE:To investigate the protective effect of PPARgamma ligands rosiglitazone on myocardium in diabetic cardiomyopathy of rats.METHODS:The rat model of diabetes was induced by administration of streptozotocin (STZ) for 6 or 10 weeks. In the treatment group the STZ-induced diabetic rats were treated with rosiglitazone. The left ventricular muscle specimens were taken from treatment and control group; then were examined under transmission electron microscope.RESULT:Cardiac myofibrils of diabetic rats in control group were obviously fewer and broken. There were fewer and smaller dissolved mitochondria with incomplete membrane and mixed cristae and karyopyknosis. Myocardium of diabetic rats treated with rosiglitazone was markedly improved although their blood glucose levels were still high.CONCLUSION:Hyperglycemia can cause destruction of myocardial cell structure. Rosiglitazone has protective effect on myocardial cells of diabetic rats, which seems to be independent of blood glucose levels.