Yuan Liu
Yuan Liu has a total of 57 co-authors. Here are the top co-authors by number of shared publications:
- Jianming Cai
Department of Radiology, The Fifth Medical Center of Chinese PLA General Hospital, Beijing, China
5 shared publications - Jinghui Dong
Department of Radiology, The Fifth Medical Center of Chinese PLA General Hospital, Beijing 100039, China
4 shared publications - Xu Bai
Department of Radiology, First Medical Center Chinese PLA General Hospital Beijing China
4 shared publications - Changchun Liu
Department of Radiology, The Fifth Medical Center of Chinese PLA General Hospital, Beijing, China
3 shared publications - Di Zhao
Senior Department of Hepatology, The Fifth Medical Center of Chinese PLA General Hospital, Beijing 100039, China
2 shared publications - Fugeng Sheng
Department of Radiology The Fifth Medical Center of Chinese PLA General Hospital Beijing China
2 shared publications
Latest Publications
- 9 September 2026
- Published by Wiley in Cochrane Database of Systematic Reviews
- Vol. 2026 (9)
Imaging‑based models for predicting cerebrovascular complications of carotid stenosis
- 9 September 2026
- Published by Wiley in Cochrane Database of Systematic Reviews
- Vol. 2026 (9)
Imaging‑based models for predicting cerebrovascular complications of carotid stenosis
- Open Access
- 12 August 2024
- Published by Elsevier in Journal of Clinical and Experimental Hepatology
- Vol. 15 (1) , 102403
Prediction Modelling for Gastroesophageal Variceal Bleeding in Patients With Chronic Hepatitis B Using Four-dimensional Flow MRI
- 12 August 2024
- Published by Elsevier in Journal of Clinical and Experimental Hepatology
- Vol. 15 (1) , 102403
Open AccessPrediction Modelling for Gastroesophageal Variceal Bleeding in Patients With Chronic Hepatitis B Using Four-dimensional Flow MRI
- Open Access
- 1 September 2023
- Published by Wolters Kluwer Health in Journal of Computer Assisted Tomography
- Vol. 47 (5) , 729-737
Predicting Breast Cancer Subtypes Using Magnetic Resonance Imaging Based Radiomics With Automatic Segmentation
Objective: The aim of the study is to demonstrate whether radiomics based on an automatic segmentation method is feasible for predicting molecular subtypes. Methods: This retrospective study included 516 patients with confirmed breast cancer. An automatic segmentation—3-dimensional UNet-based Convolutional Neural Networks, trained on our in-house data set—was applied to segment the regions of interest. A set of 1316 radiomics features per region of interest was extracted. Eighteen cross-combination radiomics methods—with 6 feature selection methods and 3 classifiers—were used for model selection. Model classification performance was assessed using the area under the receiver operating characteristic curve (AUC), accuracy, sensitivity, and specificity. Results: The average dice similarity coefficient value of the automatic segmentation was 0.89. The radiomics models were predictive of 4 molecular subtypes with the best average: AUC = 0.8623, accuracy = 0.6596, sensitivity = 0.6383, and specificity = 0.8775. For luminal versus nonluminal subtypes, AUC = 0.8788 (95% confidence interval [CI], 0.8505–0.9071), accuracy = 0.7756, sensitivity = 0.7973, and specificity = 0.7466. For human epidermal growth factor receptor 2 (HER2)–enriched versus non-HER2–enriched subtypes, AUC = 0.8676 (95% CI, 0.8370–0.8982), accuracy = 0.7737, sensitivity = 0.8859, and specificity = 0.7283. For triple-negative breast cancer versus non–triple-negative breast cancer subtypes, AUC = 0.9335 (95% CI, 0.9027–0.9643), accuracy = 0.9110, sensitivity = 0.4444, and specificity = 0.9865. Conclusions: Radiomics based on automatic segmentation of magnetic resonance imaging can predict breast cancer of 4 molecular subtypes noninvasively and is potentially applicable in large samples.
- 1 September 2023
- Published by Wolters Kluwer Health in Journal of Computer Assisted Tomography
- Vol. 47 (5) , 729-737
Open AccessPredicting Breast Cancer Subtypes Using Magnetic Resonance Imaging Based Radiomics With Automatic Segmentation
Objective: The aim of the study is to demonstrate whether radiomics based on an automatic segmentation method is feasible for predicting molecular subtypes. Methods: This retrospective study included 516 patients with confirmed breast cancer. An automatic segmentation—3-dimensional UNet-based Convolutional Neural Networks, trained on our in-house data set—was applied to segment the regions of interest. A set of 1316 radiomics features per region of interest was extracted. Eighteen cross-combination radiomics methods—with 6 feature selection methods and 3 classifiers—were used for model selection. Model classification performance was assessed using the area under the receiver operating characteristic curve (AUC), accuracy, sensitivity, and specificity. Results: The average dice similarity coefficient value of the automatic segmentation was 0.89. The radiomics models were predictive of 4 molecular subtypes with the best average: AUC = 0.8623, accuracy = 0.6596, sensitivity = 0.6383, and specificity = 0.8775. For luminal versus nonluminal subtypes, AUC = 0.8788 (95% confidence interval [CI], 0.8505–0.9071), accuracy = 0.7756, sensitivity = 0.7973, and specificity = 0.7466. For human epidermal growth factor receptor 2 (HER2)–enriched versus non-HER2–enriched subtypes, AUC = 0.8676 (95% CI, 0.8370–0.8982), accuracy = 0.7737, sensitivity = 0.8859, and specificity = 0.7283. For triple-negative breast cancer versus non–triple-negative breast cancer subtypes, AUC = 0.9335 (95% CI, 0.9027–0.9643), accuracy = 0.9110, sensitivity = 0.4444, and specificity = 0.9865. Conclusions: Radiomics based on automatic segmentation of magnetic resonance imaging can predict breast cancer of 4 molecular subtypes noninvasively and is potentially applicable in large samples.
- 6 October 2022
- Published by American Society of Neuroradiology (ASNR) in American Journal of Neuroradiology
- Vol. 43 (11) , 1582-1588
The Value of Preoperative Phase-Contrast MRI in Predicting the Clinical Outcome of Moyamoya Disease after Encephalo-Duro-Arterial Synangiosis Surgery
BACKGROUND AND PURPOSE: In patients with Moyamoya disease, the relationship between preoperative hemodynamic status and prognosis after encephalo-duro-arterial synangiosis (EDAS) surgery was unclear. We aimed to explore the value of the preoperative hemodynamic status acquired by cine phase-contrast MR imaging in predicting collateral formation and clinical outcomes after EDAS surgery in patients with Moyamoya disease. MATERIALS AND METHODS: Participants with Moyamoya disease were prospectively recruited and underwent preoperative phase-contrast MR imaging. All participants were classified into good and poor groups according to the collateral formation after EDAS surgery. On the basis of the change in the mRS system, participants were classified into the improved mRS group and the poor response group. Hemodynamic status including mean velocity, peak velocity, and blood volume flow of the superficial temporal artery was compared between groups. Logistic regression was performed to relate the phase-contrast MR imaging parameters to collateral formation and clinical outcomes. RESULTS: A total of 45 patients with Moyamoya disease with unilateral EDAS surgery were finally included. Mean velocity, peak velocity, and blood volume flow of the ipsilateral superficial temporal artery were significantly greater in patients with good collateral formation compared with those with poor collateral formation (P = .011, .004, and .013, respectively). The mean velocity, peak velocity, and blood volume flow were independently associated with postoperative collateral formation after adjusting for confounding factors. Furthermore, the peak velocity of the ipsilateral superficial temporal artery was also significantly associated with improvement of the mRS score. CONCLUSIONS: Good hemodynamic status of the ipsilateral superficial temporal artery as a donor artery evaluated by phase-contrast MR imaging was significantly associated with better collateral formation and improved mRS after EDAS surgery in patients with Moyamoya disease.
- 6 October 2022
- Published by American Society of Neuroradiology (ASNR) in American Journal of Neuroradiology
- Vol. 43 (11) , 1582-1588
The Value of Preoperative Phase-Contrast MRI in Predicting the Clinical Outcome of Moyamoya Disease after Encephalo-Duro-Arterial Synangiosis Surgery
BACKGROUND AND PURPOSE: In patients with Moyamoya disease, the relationship between preoperative hemodynamic status and prognosis after encephalo-duro-arterial synangiosis (EDAS) surgery was unclear. We aimed to explore the value of the preoperative hemodynamic status acquired by cine phase-contrast MR imaging in predicting collateral formation and clinical outcomes after EDAS surgery in patients with Moyamoya disease. MATERIALS AND METHODS: Participants with Moyamoya disease were prospectively recruited and underwent preoperative phase-contrast MR imaging. All participants were classified into good and poor groups according to the collateral formation after EDAS surgery. On the basis of the change in the mRS system, participants were classified into the improved mRS group and the poor response group. Hemodynamic status including mean velocity, peak velocity, and blood volume flow of the superficial temporal artery was compared between groups. Logistic regression was performed to relate the phase-contrast MR imaging parameters to collateral formation and clinical outcomes. RESULTS: A total of 45 patients with Moyamoya disease with unilateral EDAS surgery were finally included. Mean velocity, peak velocity, and blood volume flow of the ipsilateral superficial temporal artery were significantly greater in patients with good collateral formation compared with those with poor collateral formation (P = .011, .004, and .013, respectively). The mean velocity, peak velocity, and blood volume flow were independently associated with postoperative collateral formation after adjusting for confounding factors. Furthermore, the peak velocity of the ipsilateral superficial temporal artery was also significantly associated with improvement of the mRS score. CONCLUSIONS: Good hemodynamic status of the ipsilateral superficial temporal artery as a donor artery evaluated by phase-contrast MR imaging was significantly associated with better collateral formation and improved mRS after EDAS surgery in patients with Moyamoya disease.
- 9 August 2022
- Published by Springer Nature in Abdominal Radiology
- Vol. 47 (11) , 3712-3723
Limited diagnostic value of liver stiffness for clinically significant portal hypertension in HBV-related cirrhosis
Purpose Hepatic venous pressure gradient (HVPG) is the gold standard for portal pressure in cirrhosis, but most previous studies focused on the diagnostic value of clinically significant portal hypertension (CSPH) based on the correlation between liver stiffness (LS) and HVPG in hepatitis C virus (HCV) patients and alcoholic liver. Therefore, it is necessary to clarify the diagnostic value of LS for CSPH and the correlation with HVPG in hepatitis B virus (HBV) patients. Methods A total of 137 patients from the Fifth Medical Center of PLA General Hospital were divided into HBV group and non-HBV group according to etiology. Correlation analysis and ROC were used to analyze the correlation between LS and HVPG and the diagnostic value of CSPH. Results There was a good correlation between LS and HVPG in the total cohort and non-HBV cohort (r = 0.398, P < 0.001; r = 0.575, P < 0.001, respectively). However, the correlation between LS and HVPG was acceptable in the HBV cohort (r = 0.316, P = 0.002). When adjustment for age, MELD score, and INR, the result was still the same. Similar results were observed in the prediction for CSPH. LS showed good diagnostic value for CSPH in the total cohort and non-HBV cohort (AUC = 0.732, AUC = 0.829, respectively). However, it performed poorly in the HBV cohort (AUC = 0.689). Conclusion The etiology of HBV might affect the diagnostic performance of LS for predicting CSPH. Graphical abstract
- 9 August 2022
- Published by Springer Nature in Abdominal Radiology
- Vol. 47 (11) , 3712-3723
Limited diagnostic value of liver stiffness for clinically significant portal hypertension in HBV-related cirrhosis
Purpose Hepatic venous pressure gradient (HVPG) is the gold standard for portal pressure in cirrhosis, but most previous studies focused on the diagnostic value of clinically significant portal hypertension (CSPH) based on the correlation between liver stiffness (LS) and HVPG in hepatitis C virus (HCV) patients and alcoholic liver. Therefore, it is necessary to clarify the diagnostic value of LS for CSPH and the correlation with HVPG in hepatitis B virus (HBV) patients. Methods A total of 137 patients from the Fifth Medical Center of PLA General Hospital were divided into HBV group and non-HBV group according to etiology. Correlation analysis and ROC were used to analyze the correlation between LS and HVPG and the diagnostic value of CSPH. Results There was a good correlation between LS and HVPG in the total cohort and non-HBV cohort (r = 0.398, P < 0.001; r = 0.575, P < 0.001, respectively). However, the correlation between LS and HVPG was acceptable in the HBV cohort (r = 0.316, P = 0.002). When adjustment for age, MELD score, and INR, the result was still the same. Similar results were observed in the prediction for CSPH. LS showed good diagnostic value for CSPH in the total cohort and non-HBV cohort (AUC = 0.732, AUC = 0.829, respectively). However, it performed poorly in the HBV cohort (AUC = 0.689). Conclusion The etiology of HBV might affect the diagnostic performance of LS for predicting CSPH. Graphical abstract
- 1 November 2021
- Published by Bentham Science Publishers Ltd. in Current Medical Imaging
- Vol. 17 (11) , 1363-1368
Diagnostic Value of Portal Vein Velocity for Portal Hypertension in Patients with Hepatitis B Virus-related Cirrhosis
Background: Portal vein velocity (PVV) has shown reasonable correlation with the presence of portal hypertension in patients with cirrhosis. This study aims to evaluate the value of PVV for diagnosing clinically significant portal hypertension (CSPH) and predicting the risk of variceal hemorrhage (VH) in patients with hepatitis B virus (HBV)-related cirrhosis. Material and Methods: A cohort of 166 consecutive adult patients with HBV-related cirrhosis was recruited in this retrospective study from two high-volume liver centers in China between April 2015 and April 2017. The performance of PVV and other non-invasive parameters for diagnosing CSPH and predicting risk of VH were studied. Results: PVV demonstrated the best performance for diagnosing CSPH (defined as an HVPG ≥10 mmHg) in patients with HBV-related cirrhosis among the included noninvasive predictors with the area under the receiver operating characteristic curve (AUC), specificity, and sensitivity of 0.745, 50%, and 93.5%, respectively. Other noninvasive markers, including APRI, AAR, LS, FIB-4, and diameter of portal vein, did not show sufficient performance with the AUCs of 0.565, 0.560, 0.544, 0.529, and 0.474, respectively. With regard to predicting the risk of VH (defined as an HVPG ≥12 mmHg), PPV also exhibited a moderate performance with an AUC of 0.762, which was superior to that of the aforementioned markers. By using two cutoff values of PVV to rule-out (11.65 cm/s) and rule-in (20.20 cm/s) CSPH, 30 (33.7%) patients showed definite results categories, with 23 (76.7%) patients were well classified and 7 (23.3%) were misclassified. Fifty-nine (66.3%) patients were with indeterminate results. By using PVV values of 13.10 cm/s and 21.40 cm/s to rule-out and rule-in HVPG ≥ 12mmHg, 34 (38.2%) patients has definite results, among whom 26 (76.5%) were well classified and 8 (23.5%) were misclassified. And 55 (61.8%) patients required further evaluation. Conclusion: PPV is not good enough to serve as a non-invasive parameter for identifying CSPH and predicting risk of VH in patients with HBV-related cirrhosis.
- 1 November 2021
- Published by Bentham Science Publishers Ltd. in Current Medical Imaging
- Vol. 17 (11) , 1363-1368
Diagnostic Value of Portal Vein Velocity for Portal Hypertension in Patients with Hepatitis B Virus-related Cirrhosis
Background: Portal vein velocity (PVV) has shown reasonable correlation with the presence of portal hypertension in patients with cirrhosis. This study aims to evaluate the value of PVV for diagnosing clinically significant portal hypertension (CSPH) and predicting the risk of variceal hemorrhage (VH) in patients with hepatitis B virus (HBV)-related cirrhosis. Material and Methods: A cohort of 166 consecutive adult patients with HBV-related cirrhosis was recruited in this retrospective study from two high-volume liver centers in China between April 2015 and April 2017. The performance of PVV and other non-invasive parameters for diagnosing CSPH and predicting risk of VH were studied. Results: PVV demonstrated the best performance for diagnosing CSPH (defined as an HVPG ≥10 mmHg) in patients with HBV-related cirrhosis among the included noninvasive predictors with the area under the receiver operating characteristic curve (AUC), specificity, and sensitivity of 0.745, 50%, and 93.5%, respectively. Other noninvasive markers, including APRI, AAR, LS, FIB-4, and diameter of portal vein, did not show sufficient performance with the AUCs of 0.565, 0.560, 0.544, 0.529, and 0.474, respectively. With regard to predicting the risk of VH (defined as an HVPG ≥12 mmHg), PPV also exhibited a moderate performance with an AUC of 0.762, which was superior to that of the aforementioned markers. By using two cutoff values of PVV to rule-out (11.65 cm/s) and rule-in (20.20 cm/s) CSPH, 30 (33.7%) patients showed definite results categories, with 23 (76.7%) patients were well classified and 7 (23.3%) were misclassified. Fifty-nine (66.3%) patients were with indeterminate results. By using PVV values of 13.10 cm/s and 21.40 cm/s to rule-out and rule-in HVPG ≥ 12mmHg, 34 (38.2%) patients has definite results, among whom 26 (76.5%) were well classified and 8 (23.5%) were misclassified. And 55 (61.8%) patients required further evaluation. Conclusion: PPV is not good enough to serve as a non-invasive parameter for identifying CSPH and predicting risk of VH in patients with HBV-related cirrhosis.
- 1 November 2020
- Published by Elsevier in European Journal of Radiology
- Vol. 132, 109288
Budd-Chiari Syndrome: The “inferior vena cava reverse-flow” sign and “jet-blood” sign on CT and MRI
- 1 November 2020
- Published by Elsevier in European Journal of Radiology
- Vol. 132, 109288
Budd-Chiari Syndrome: The “inferior vena cava reverse-flow” sign and “jet-blood” sign on CT and MRI
- Open Access
- 18 November 2016
- Published by S. Karger AG in Cellular Physiology and Biochemistry
- Vol. 40 (1-2) , 77-90
Inhibition of STAT Pathway Impairs Anti-Hepatitis C Virus Effect of Interferon Alpha
Background/Aims: Signal transducer and activator of transcription (STAT) pathway plays an important role in antiviral efficacy of interferon alpha (IFN-α). IFN-α is the main therapeutic against hepatitis C virus (HCV) infection. We explored effects of IFN-α on HCV replication and antiviral gene expression by targeting STAT. Methods: In response to IFN-α, STAT status, HCV replication, and antiviral gene expression were analyzed in human hepatoma Huh7.5.1 cells before and after cell culture-derived HCV infection. Results: IFN-α treatment induced expression and phosphorylation of STAT1 and STAT2 in Huh7.5.1 cells. Pretreatment of Huh7.5.1 cells with a mAb to IFN alpha receptor (IFNAR) 2 decreased IFN-α-dependent phosphorylation of STAT1 and STAT2, whereas pretreatment with an IFNAR1 mAb increased such phosphorylation, suggesting that IFNAR mediates IFN-α-triggered STAT signaling. During HCV infection, STAT1 and STAT2 phosphorylation could be rescued by IFN-α and IFN-α-induced phosphorylation of STAT1 and STAT2 was impaired. Inhibition of STAT pathway by Jak inhibitor I significantly enhanced HCV RNA replication and viral protein expression. Antiviral genes coding for IFN regulatory factor 9 and IFN-stimulated gene 15 were up-regulated by IFN-α during HCV infection but such up-regulation was abrogated by Jak inhibitor I. Conclusion: These results establish that activation of STAT pathway is essential for anti-HCV efficacy of IFN-α. Impairment of IFN-α-triggered STAT signaling by HCV may account for evading IFN-α response.
- 18 November 2016
- Published by S. Karger AG in Cellular Physiology and Biochemistry
- Vol. 40 (1-2) , 77-90
Open AccessInhibition of STAT Pathway Impairs Anti-Hepatitis C Virus Effect of Interferon Alpha
Background/Aims: Signal transducer and activator of transcription (STAT) pathway plays an important role in antiviral efficacy of interferon alpha (IFN-α). IFN-α is the main therapeutic against hepatitis C virus (HCV) infection. We explored effects of IFN-α on HCV replication and antiviral gene expression by targeting STAT. Methods: In response to IFN-α, STAT status, HCV replication, and antiviral gene expression were analyzed in human hepatoma Huh7.5.1 cells before and after cell culture-derived HCV infection. Results: IFN-α treatment induced expression and phosphorylation of STAT1 and STAT2 in Huh7.5.1 cells. Pretreatment of Huh7.5.1 cells with a mAb to IFN alpha receptor (IFNAR) 2 decreased IFN-α-dependent phosphorylation of STAT1 and STAT2, whereas pretreatment with an IFNAR1 mAb increased such phosphorylation, suggesting that IFNAR mediates IFN-α-triggered STAT signaling. During HCV infection, STAT1 and STAT2 phosphorylation could be rescued by IFN-α and IFN-α-induced phosphorylation of STAT1 and STAT2 was impaired. Inhibition of STAT pathway by Jak inhibitor I significantly enhanced HCV RNA replication and viral protein expression. Antiviral genes coding for IFN regulatory factor 9 and IFN-stimulated gene 15 were up-regulated by IFN-α during HCV infection but such up-regulation was abrogated by Jak inhibitor I. Conclusion: These results establish that activation of STAT pathway is essential for anti-HCV efficacy of IFN-α. Impairment of IFN-α-triggered STAT signaling by HCV may account for evading IFN-α response.