LXLi-Yun Xue

Li-Yun Xue

*This is an algorithmically generated author record based on publicly available information.

Current Organizations / Institutes

The organizations/institutes from the scholar's latest article


Li-Yun Xue has a total of 24 co-authors. Here are the top co-authors by number of shared publications:

  • Bei-Jian Huang

    Department of Ultrasound Zhongshan Hospital Fudan University Shanghai China

    8 shared publications
  • Qing Lu

    Department of Ultrasound, Zhongshan Hospital, Fudan University, Shanghai, China

    7 shared publications
  • Wen-Ping Wang

    Department of Ultrasound, Zhongshan Hospital, Fudan University, No. 180 Fenglin Road, Xuhui District, Shanghai, 200032, China

    6 shared publications
  • Jie-Xian Wen

    Department of Ultrasound, Zhongshan Hospital of Fudan University, No. 180 of Fenglin Road, Xuhui District, Shanghai 200032, China; Shanghai Institute of Medical Imaging, Fudan University, No. 180 of Fenglin Road, Shanghai 200032, China

    5 shared publications
  • Cui-Xian Li

    Department of Ultrasound, Zhongshan Hospital, Fudan University, Shanghai, China

    4 shared publications
  • Li-Xia Yan

    Department of Ultrasound, Zhongshan Hospital, Fudan University, Shanghai, China

    4 shared publications

Latest Publications

  • Open Access

    Transfer learning radiomics based on multimodal ultrasound imaging for staging liver fibrosis

    by Li-Yun Xue, Zhuo-Yun Jiang, Tian-Tian Fu, Qing-Min Wang, Yu-Li Zhu, Meng Dai, Wen-Ping Wang, Jin-Hua Yu, Hong Ding

    Objectives: To propose a transfer learning (TL) radiomics model that efficiently combines the information from gray scale and elastogram ultrasound images for accurate liver fibrosis grading. Methods: Totally 466 patients undergoing partial hepatectomy were enrolled, including 401 with chronic hepatitis B and 65 without fibrosis pathologically. All patients received elastography and got liver stiffness measurement (LSM) 2–3 days before surgery. We proposed a deep convolutional neural network by TL to analyze images of gray scale modality (GM) and elastogram modality (EM). The TL process was used for liver fibrosis classification by Inception-V3 network which pretrained on ImageNet. The diagnostic performance of TL and non-TL was compared. The value of single modalities, including GM and EM alone, and multimodalities, including GM + LSM and GM + EM, was evaluated and compared with that of LSM and serological indexes. Receiver operating characteristic curve analysis was performed to calculate the optimal area under the curve (AUC) for classifying fibrosis of S4, ≥ S3, and ≥ S2. Results: TL in GM and EM demonstrated higher diagnostic accuracy than non-TL, with significantly higher AUCs (all p < .01). Single-modal GM and EM both performed better than LSM and serum indexes (all p < .001). Multimodal GM + EM was the most accurate prediction model (AUCs are 0.950, 0.932, and 0.930 for classifying S4, ≥ S3, and ≥ S2, respectively) compared with GM + LSM, GM and EM alone, LSM, and biomarkers (all p < .05). Conclusions: Liver fibrosis can be staged by a transfer learning modal based on the combination of gray scale and elastogram ultrasound images, with excellent performance. Key Points: • Transfer learning consists in applying to a specific deep learning algorithm that pretrained on another relevant problem, expected to reduce the risk of overfitting due to insufficient medical images. • Liver fibrosis can be staged by transfer learning radiomics with excellent performance. • The most accurate prediction model of transfer learning by Inception-V3 network is the combination of gray scale and elastogram ultrasound images.


  • Differentiation of subtypes of renal cell carcinoma with contrast-enhanced ultrasonography

    by Li-Yun Xue, Qing Lu, Bei-Jian Huang, Cui-Xian Li, Li-Xia Yan, Wen-Ping Wang

    We aimed to assess the difference of enhancement patterns among the three RCC subtypes with contrast-enhanced ultrasound (CEUS). Two hundreds cases of pathologically proved clear cell renal cell carcinomas (ccRCC), 58 papillary renal cell carcinomas


  • Quantitative evaluation of contrast-enhanced ultrasound for differentiation of renal cell carcinoma subtypes and angiomyolipoma

    by Cui-Xian Li, Qing Lu, Bei-Jian Huang, Li-Yun Xue, Li-Xia Yan, Feng-Yang Zheng, Jie-Xian Wen, Wen-Ping Wang


  • Differentiation of Renal Tumor Histotypes: Usefulness of Quantitative Analysis of Contrast-Enhanced Ultrasound

    by Qing Lu, Bei-Jian Huang, Li-Yun Xue, Pei-Li Fan, Wen-Ping Wang

    OBJECTIVE. The purpose of this study is to evaluate quantitative analysis of contrast-enhanced ultrasound (CEUS) in the differential diagnosis of renal tumor histotypes. MATERIALS AND METHODS. Between January 2010 and December 2013, 106 clear cell renal cell carcinomas (RCCs) (mean [± SD] diameter, 3.7 ± 1.8 cm), 34 angiomyolipomas (mean diameter, 4.1 ± 1.4 cm), 25 papillary RCCs (mean diameter, 3.5 ± 1.1 cm), and 28 chromophobe RCCs (mean diameter, 2.9 ± 0.9 cm) underwent CEUS quantitative analysis. The dynamic vascular pattern was analyzed with the Fisher exact chi-square test, and rise time, time to peak (TTP), and tumor-to-cortex enhancement ratio were analyzed with the independent-sample t test. RESULTS. Dynamic vascular pattern types I and III (hyperenhancement) were more common among clear cell RCCs, whereas type II (hypoenhancement) was more common among angiomyolipomas, papillary RCCs, and chromophobe RCCs. Irrespective of dynamic vascular pattern class, the rise time and TTP were the shortest in...


  • Dynamic enhancement pattern of intrahepatic cholangiocarcinoma on contrast-enhanced ultrasound: the correlation with cirrhosis and tumor size

    by Qing Lu, Li-Yun Xue, Wen-Ping Wang, Bei-Jian Huang, Cui-Xian Li

    To retrospectively evaluate the dynamic enhancement pattern of contrast-enhanced ultrasound (CEUS) in intrahepatic cholangiocarcinoma (ICC) of varying sizes and hepatic backgrounds. CEUS was performed in 98 pathologically confirmed ICCs (n = 39 < 30 mm, n = 59 > 30 mm; n = 45 with cirrhosis and n = 53 with normal liver). The dynamic enhancement pattern of CEUS was retrospectively analyzed. In the arterial phase, heterogeneous hyper-enhancement was more frequent in ICCs with cirrhosis (21/45, 46.7% vs. 11/53, 20.8% in ICCs with normal liver, p = 0.009), while peripheral hyper-enhancement and hypo-enhancement were more common in ICCs with normal liver (14/53, 26.4%; 11/53, 20.8% vs. 2/45, 4.4%; 2/45, 4.4% in ICCs with cirrhosis, p = 0.005 and 0.033, respectively). There were no significant differences between portal and delayed phases. In ICCs < 30 mm, homogeneous hyper-enhancement was more frequently identified (27/39, 69.2% vs. 10/59, 16.9% in ICCs > 30 mm, p < 0.001), whereas in ICCs > 30 mm, heterogeneous, and peripheral hyper-enhancement were more commonly observed (26/59, 44.1% vs. 6/39, 15.4% in ICCs < 30 mm, p = 0.004, and 14/59, 23.7% vs. 2/39, 5.1% in ICCs < 30 mm, p = 0.023, respectively). The washout pattern in portal and delayed phases were not significantly different in ICCs with different sizes. 60.7% (17/28) ICCs < 30 mm and 85.2% (23/27) ICCs > 30 mm with cirrhosis, together with 66.7% (14/21) ICCs < 30 mm with normal liver displayed intense contrast agent uptake (homogeneous or heterogeneous hyper-enhancement) in arterial phase followed by washout in portal and delayed phase, which was much higher than that in ICCs > 30 mm with normal liver (34.4%, 11/32, p < 0.001, <0.001 and =0.027, respectively). The CEUS dynamic enhancement pattern of ICC varies with size and hepatic background. The enhancement pattern is indistinguishable from hepatocellular carcinoma on CEUS in most ICCs with cirrhosis and in most ICCs < 30 mm with normal liver.


  • Histotype differentiation of hypo-echoic renal tumors on CEUS: usefulness of enhancement homogeneity and intensity

    by Qing Lu, Li-Yun Xue, Bei-Jian Huang, Wen-Ping Wang, Cui-Xian Li

    The purpose of this study is to evaluate qualitative and quantitative analysis of contrast-enhanced ultrasound (CEUS) in differential diagnoses of hypo-echoic renal tumor histotypes.


    • 1 October 2014
      • Vol. 18  (99) , 179-88

    The value of contrast-enhanced ultrasound (CEUS) in detecting minute renal cell carcinoma.

    by Cui-Xian Li, Qing Lu, Bei-Jian Huang, Li-Yun Xue, Li-Xia Yan, Jie-Xian Wen, Wen-Ping Wang

    To investigate the value of contrast-enhanced ultrasound (CEUS) in the detection of minute renal cell carcinoma (MRCC) compared to conventional ultrasound (US) and contrast-enhanced computed tomography (CECT). Thirty-eight consecutive patients with 38 histopathologically proven MRCCs (≦15 mm) were enrolled in our study. CEUS and CECT were available in 38 and 24 patients, respectively. The features of CEUS were evaluated and compared to conventional ultrasound (US) and CECT. Ten (26.3%) tumors could not be detected by conventional US, while all tumors were detected by CEUS. The features of tumor border, blood flow, and echogenicity had significant difference between conventional US and CEUS (p=0.000, p=0.003, and p=0.012, respectively). The score of visibility of tumors by CEUS was significant higher than that of conventional US. The sensitivity, specificity, and accuracy of conventional US and CEUS in evaluating tumor necrosis were 42.9%, 50%, and 47.4% vs. 85.7%, 95.8%, and 92.1%, respectively. The enhancement features of MRCC including tumor vascularization, homogeneity, and border had no significant difference between CEUS and CECT (all p>0.05). On CEUS, synchronous-in (89.5%), hypervascular (84.2%), and fast-out (71.1%) were the most commonly observed enhancement characteristics for MRCC. CEUS performs better in detecting MRCC than conventional US, and it has the same capabilities in reflecting the enhanced features of MRCC as CECT.


  • Triphasic and epithelioid minimal fat renal angiomyolipoma and clear cell renal cell carcinoma: qualitative and quantitative CEUS characteristics and distinguishing features

    by Qing Lu, Cui-Xian Li, Bei-Jian Huang, Li-Yun Xue, Wen-Ping Wang

    To determine the contrast-enhanced ultrasonography (CEUS) characteristics of minimal fat renal angiomyolipoma (AML) (triphasic and epithelioid) and compare them to each other and to clear cell renal cell carcinoma (ccRCC) to explore their differential diagnostic clue. Qualitative and quantitative CEUS analyses were retrospectively conducted for epithelioid renal AMLs (EAMLs) (n = 15), triphasic minimal fat AMLs (TAMLs) (n = 25), and ccRCCs (n = 113). Enhancement patterns and features with CEUS were qualitatively evaluated. As for the quantitative parameters, rise times (RT), time to peak (TTP), and tumor-to-cortex enhancement ratio (TOC ratio) were compared among these renal tumor histotypes. No significant differences were detected on conventional ultrasound in the three histotypes of renal tumor. On qualitative CEUS analysis, centripetal enhancement in cortical phase (73.3% in EAMLs, 84.0% in TAMLs vs. 18.6% in ccRCCs, p p p = 0.034 and 0.013, respectively) were valuable traits for differentiating EAMLs and TAMLs from ccRCCs. Furthermore, with quantitative analysis, RT and TTP were much shorter in ccRCCs than those in EAMLs and TAMLs. However, all these qualitative and quantitative characteristics made no significant difference between EAMLs and TAMLs. In the differential diagnosis of EAMLs from TAMLs, pseudocapsule sign was valuable (40.0% in EAMLs vs. 0.0% in TAMLs, p p 97.34% as the criteria to differentiate ccRCCs and EAMLs from TAMLs, the sensitivity and specificity were 80.0% and 87.5%, respectively. Qualitative and quantitative CEUS analyses are helpful in the differential diagnosis of ccRCCs, EAMLs, and TAMLs.


  • Contrast-enhanced ultrasonography for evaluation of cystic renal mass: in comparison to contrast-enhanced CT and conventional ultrasound

    by Li-Yun Xue, Qing Lu, Bei-Jian Huang, Jiao-Jiao Ma, Li-Xia Yan, Jie-Xian Wen, Wen-Ping Wang

    To assess the value of contrast-enhanced ultrasonography (CEUS) in evaluating cystic renal lesions compared with conventional ultrasound (US) and contrast-enhanced computed tomography (CECT). One hundred and three patients with complex cystic renal masses underwent preoperative US and CEUS, among which 70 conducted CECT at our institution. The images were analyzed with the number of septa, septa and wall thickness and the presence of solid component, and final diagnosis was made. In malignancies, CEUS demonstrated more septa, thicker wall or septa, and more solid components than US and CECT. CEUS permitted categorization of 51.7% (30/58) and 28.6% (10/35) of malignant tumors in higher grade than by US and CECT, respectively. In benign lesions, CEUS detected more septa than CECT and correctly diagnosed benign cysts which appeared as solid lesions in US. CEUS permitted downgrading of 71.1% (32/45) and 17.1% (6/35) of benign lesions compared to US and CECT. The diagnostic performance of CEUS was better than US for benign cystic lesions. The phenomenon that solid-like component by US did not enhance by CEUS was a strong predictor of benign disease, with a positive predictive value (PPV) of 100%. Enhancement of solid, soft tissue by CEUS was highly predictive of malignancy, with a PPV of 100%. CEUS was superior to US and CECT in visualizing the number of septa septa and wall thickness, and the presence of solid component of cystic renal lesions. CEUS may play a similar role to CECT in the diagnosis of renal cystic lesions, and better than US.