SFShu-Feng Fan

Shu-Feng Fan

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

Current Organizations / Institutes

The organizations/institutes from the scholar's latest article

Previous Organizations / Institutes

The organizations / institutes from the scholar's other articles in the past ten years


Shu-Feng Fan has a total of 98 co-authors. Here are the top co-authors by number of shared publications:


Latest Publications

  • Open Access

    Native T1 mapping radiomics for assessing FIB-4-associated fibrosis risk in metabolic dysfunction-associated steatotic liver disease

    by Jiaqi Duan, Tianhong Yin, Shiyi Xu, Guoyang Wang, Yihua Wang, Qinpan Rao, Xiaoshan Huang, Xiaozhong Zheng, Jianxia Xu, Shufeng Fan

    Background: The fibrosis-4 index (FIB-4) is widely used for noninvasive fibrosis risk stratification in patients with metabolic dysfunction-associated steatotic liver disease (MASLD), but it does not directly characterize spatial heterogeneity within the hepatic parenchyma. This study aimed to investigate whether native T1 mapping radiomics features are associated with continuous FIB-4 values in patients with MASLD, after accounting for hepatic steatosis and iron deposition with the goal of enabling early identification of liver fibrosis. Methods: This retrospective single-center study included 124 patients with MASLD who underwent multiparametric liver magnetic resonance imaging, including proton density fat fraction (PDFF), iron (R2*), and native T1 mapping. To prevent data leakage, a 5-fold cross-validation framework was employed for radiomics analysis; variance thresholding, univariate Spearman correlation, and elastic net regression with nested cross-validation were performed sequentially within each training fold for feature selection and Z-score standardization. FIB-4 was treated as a continuous biochemical index, and two hierarchical regression models were refitted on the full dataset to evaluate the incremental value of the selected native T1 radiomics features beyond baseline liver fat (PDFF) and iron (R2*) content (Model 1: baseline; Model 2: baseline + radiomics features). In addition, we conducted an age-adjusted sensitivity analysis by including age as an additional covariate in both models to examine whether the contribution of radiomics features persists after accounting for age-related imaging variability. Results: Twelve robust radiomics features were retained through the cross-validation selection pipeline. Model 1 showed limited explanatory ability for FIB-4 values (R2=0.028, adjusted R2=0.012, P=0.179). After incorporating the selected native T1 radiomics features, Model 2 demonstrated a significantly improved model fit (R2=0.455, adjusted R2=0.376; ΔR2=0.427; F-change =6.004, P2*, the inclusion of radiomics features still significantly increased the explained variance, raising R2 from 0.428 to 0.631 (ΔR2=0.203, PConclusions: Using FIB-4 as a clinical reference, and after excluding the effects of PDFF and R2*, a significant within-sample correlation was observed between the radiomics features generated by native T1 mapping and FIB-4 values, indicating its potential clinical utility in the noninvasive assessment of fibrosis-related biochemical risk trends in patients with MASLD.

  • Open Access

    Imaging features, diagnosis, and clinical management of hepatic involvement in hereditary hemorrhagic telangiectasia: a case series report and literature review

    by Qiubo Feng, Qinpan Rao, Jiaqi Duan, Shufeng Fan, Linying Ma

    To explore the clinical characteristics, imaging features, key diagnostic and therapeutic points, and risks of interventional therapy in hereditary hemorrhagic telangiectasia (HHT) with hepatic involvement, so as to improve the ability of clinicians and radiologists in the identification, diagnosis and treatment of this disease. A retrospective analysis was performed on the clinical data of 3 patients with HHT who were admitted to our hospital in the past 5 years and presented with predominant hepatic involvement. Their clinical symptoms, laboratory test results, imaging findings, and treatment outcomes were summarized, and a literature review was conducted for comprehensive analysis. All 3 patients met the Curacao diagnostic criteria for HHT and were found to have diffuse hepatic arteriovenous malformations, mainly presenting as mixed shunts. Computed tomography (CT) and magnetic resonance imaging (MRI) revealed dilated hepatic arteries, early visualization of the portal vein and hepatic vein, diffuse intrahepatic telangiectasias, and abnormal perfusion. One patient was complicated with biloma, and another was associated with portal hypertension and esophagogastric varices. Patient 3 underwent transcatheter hepatic arterial embolization (TAE) due to abdominal pain, and extensive thrombosis of the main portal vein and superior mesenteric vein occurred after the operation. Hepatic involvement in HHT is characterized by diffuse hepatic vascular malformations, with insidious clinical manifestations that are prone to misdiagnosis. Doppler ultrasound is the preferred screening method, while CT/MRI can comprehensively evaluate vascular malformations and related complications. Hepatic arterial embolization has a high risk of inducing portal vein thrombosis, and its routine application should be avoided in strict accordance with international guidelines. Clinicians need to strengthen medical history collection and multimodal imaging evaluation to reduce misdiagnosis and mistreatment.

  • Open Access

    Small intestinal metastasis of triple-negative breast cancer mimicking lymphoma: imaging findings and case report

    by Lin-Ying Ma, Xiao-Shan Huang, Shu-Feng Fan, Li-Ming Xue

    Small intestinal metastasis from breast cancer is a rare entity with non-specific clinical presentations, rendering it susceptible to missed or delayed diagnosis. Herein, we present a rare case of breast cancer metastasizing to the small bowel, with unexplained anemia as the sole initial clinical complaint.The patient manifested persistent dizziness and palpitations, accompanied by repeatedly positive fecal occult blood testing, and carried a well-documented history of surgical resection for bilateral breast malignancy.Contrast-enhanced computed tomography (CT) of the small bowel demonstrated segmental, asymmetric wall thickening with adjacent lymphadenopathy and no luminal stenosis; the thickened wall exhibited heterogeneous enhancement, raising initial suspicion of lymphoma. Subsequent double-balloon enteroscopy(DBE) identified a large circumferential deep ulcerative lesion, which was endoscopically indistinguishable from primary small bowel adenocarcinoma or intestinal lymphoma. The definitive diagnosis of small intestinal metastasis from breast cancer was established by histopathological and immunohistochemical evaluation, in conjunction with the patient’s oncological history. A review of the literature indicates that small intestinal metastasis from breast cancer is more commonly associated with invasive lobular carcinoma, and affected patients usually present during routine surveillance or with abdominal pain. In the present case, the primary pathology was invasive ductal carcinoma of triple-negative phenotype (TNBC), and the patient’s initial presentation with anaemia constitutes an atypical manifestation. Clinicians should maintain a high index of suspicion for small bowel metastasis in breast cancer survivors who develop unexplained anemia, positive fecal occult blood, or abnormal small intestinal imaging, even in the absence of classic gastrointestinal symptoms. A multidisciplinary approach integrating imaging, endoscopy, pathology, and immunohistochemistry is essential to minimise diagnostic delay and avoid misdiagnosis.

  • Open Access

    A CT-based decision tree model for differentiating sub-3 cm gastric ectopic pancreas from gastrointestinal stromal tumors

    by Jiaqi Duan, Yunzhi Zhao, Qihao Shi, Yining Wang, Xiaoshan Huang, Qinpan Rao, Shufeng Fan, Jianxia Xu

    To develop a CT-based decision tree model integrating clinical and imaging features for the preoperative differentiation of gastric ectopic pancreas (GEPs) and gastrointestinal stromal tumors (GISTs) with a maximum diameter of less than 3 cm. This retrospective study included 86 patients with pathologically confirmed GEPs (n = 26) and GISTs (n = 60), all with lesions smaller than 3 cm. Clinical information and CT features were collected. The dataset was divided into training and testing sets. A decision tree classification model was constructed using key variables selected from the training set via univariate analyses and logistic regression. The decision tree's hyperparameters were optimised using five-fold cross-validation. Diagnostic performance was evaluated on an independent test set, including plotting ROC curves to calculate AUC values, sensitivity, and specificity, alongside using calibration curves to assess goodness-of-fit. Furthermore, the SHAP method was employed to provide visual explanations for the final model's predictions. The decision tree model identified four key variables: age (clinical factor) and three CT features: ratio of lesion-to-pancreas attenuation in the arterial phase(A2), lesion long-to-short diameter ratio (LD/SD ratio), and intralesional low attenuation (ILA). The model, based on these four features, achieved an AUC of 0.744(95% CI:0.589-0.950), with sensitivity of 76.9% and specificity of 84.6%. Concurrently, calibration analysis substantiated the model's exceptional predictive precision. The Brier score (0.0648) and the Hosmer-Lemeshow test (χ2 = 5.365, df = 8, P = 0.718) both demonstrated a high degree of agreement between the model's predicted probabilities and the actual observed values. The CT-based decision tree model, integrating four clinical and CT features, provides a reliable and visualized tool for differentiating GEPs from GISTs with a maximum diameter of less than 3 cm, demonstrating strong diagnostic performance.

  • Open Access

    Disseminated cryptococcosis involving the adrenal gland and brain in an immunocompetent patient: A case report

    by Qinpan Rao, Fang Wu, Dongying Su, Xia Song, Jinzhan Su, Linying Ma, Jianxia Xu, Shufeng Fan, Jinwang Zhu, Jianmin Wen

    Rationale: Cryptococcosis is a fungal infection that primarily occurs in immunocompromised individuals. This infection typically affects the lungs and central nervous system, while adrenal involvement is relatively uncommon. Our research reports a rare case of adrenal cryptococcosis in a patient with normal immune function, with suspected involvement of the nervous system. Patient concerns: A 51-year-old male patient with normal immune function presented with progressive symptoms of dizziness, palpitations, and fatigue, which had persisted for 2 years. Imaging studies revealed a hypoechoic mass in the left adrenal gland, multiple intracranial lesions, and progressive hydrocephalus. The adrenal mass was initially suspected to be a tumor, but pathological examination following surgical resection confirmed a Cryptococcus infection. Despite normal endocrine and immune markers, central nervous system dissemination was strongly suspected based on the progressive changes observed on magnetic resonance imaging. The patient was treated with a combination of amphotericin B and flucytosine, leading to significant improvement in the neurological symptoms. Diagnoses: The diagnosis of adrenal cryptococcosis was confirmed through surgical pathology, and based on the results of cranial magnetic resonance imaging, it is considered that the adrenal cryptococcosis may be accompanied by potential involvement of the nervous system. Interventions and outcomes: After the diagnosis, a treatment plan was developed, and following treatment, the patient’s dizziness symptoms improved, leading to their subsequent discharge. Lessons: These results suggest that clinicians should broaden their diagnostic thinking by combining advanced imaging techniques with thorough clinical evaluation to enhance their diagnostic ability for adrenal cryptococcosis, thereby preventing missed or misdiagnoses and ensuring timely, accurate, and effective treatment to improve patient outcomes.


  • Open Access

    The correlation between the styloid process length, angle, head shape, and eagle syndrome based on high-resolution CT three-dimensional reconstruction: a retrospective study

    by Jiaqi Duan, Fang Wu, Linfeng Liu, Jianxia Xu, Xiaozhong Zheng, Shufeng Fan, Qinpan Rao

    Objective: A comprehensive evaluation index (R value) based on the styloid process (SP) length and spatial angle orientation was constructed to explore the imaging and clinical features of Eagle Syndrome, and assess its diagnostic value in the identification of Eagle Syndrome.Methods: A retrospective analysis was conducted on the high-resolution CT (HRCT) three-dimensional reconstruction maximum intensity projection (MIP), with a data of 57 clinically diagnosed Eagle Syndrome (ES group) and 49 Non Eagle Syndrome (NES group). The bilateral SP length, inward angulation, and forward tilt angle were measured. A comprehensive evaluation index, the R value, was introduced based on the head shape. Independent two-sample t-tests were used to compare the differences in parameters between the two groups, and ROC curve analysis was performed to assess the diagnostic efficacy of the R value. Finally, binary logistic regression was employed to validate the stability of the model.Results: The ES Group exhibited significantly higher parameters compared to the NES Group in terms of SP length (left: 34.19 ± 5.14 mm, right: 34.13 ± 6.40 mm), inward angulation (left: 24.29° ± 3.09°, right: 22.22° ± 3.18°), and forward tilt angle (left: 28.39 ± 2.76°, right: 28.29 ± 2.72°). The ROC curve analysis of the R value showed that the left side had a AUC of 0.86 (95% CI, 0.79–0.93), with an optimal cutoff value of 2.85, sensitivity of 82.5%, and specificity of 79.6%. The right side had an AUC of 0.82 (95% CI, 0.74–0.90), with an optimal cutoff value of 2.72, sensitivity of 78.9%, and specificity of 75.5%. The binary logistic regression results demonstrated that the R value exhibits excellent discriminative ability in the diagnosis of Eagle Syndrome. In particular, when the left R > 2.85 and/or right R > 2.72, Eagle Syndrome should be strongly suspected, and a precise diagnosis should be made in combination with clinical symptoms.Conclusion: Eagle Syndrome is closely related to the length and angle of the SP. The R value, as a composite evaluation index integrating key anatomical parameters such as length and angle, demonstrates high diagnostic efficacy and significant clinical utility. Moreover, R value (The left R > 2.85 and/or right R > 2.72) can be applied as quantitative reference criteria for diagnosis.


  • Open Access

    Computed tomography manifestations of drug-induced liver injury according to type and severity of injury

    by Yao Chen, Yuzhen Xi, Fanfan Zhao, Huanhuan Li, Min Zhou, Yue Xu, Shufeng Fan, Miao Liu

    Background: Drug-induced liver injury (DILI) has become a major cause of acute liver failure, and its incidence has been increasing steadily in recent years. This study aimed to compare the clinical and computed tomography (CT) imaging features of the variable biochemical damage and severity of DILI to establish a radiological model for predicting high-risk DILI based on CT image features. Methods: The eligible patients with DILI (January 2016 to March 2024) who underwent serum laboratory examination and contrast abdominal CT within 3 months of onset were retrospectively analyzed at Affiliated Xihu Hospital of Hangzhou Medical College (Institution I) and The Second Affiliated Hospital of Zhejiang Chinese Medical University (Institution II). The severity-associated CT features were determined via binomial logistic regression analysis, and the efficacy of the different models were compared. The odds ratios (ORs) and corresponding 95% confidence intervals (CIs) provided were not adjusted. Results: The injury types included hepatocellular (n=68, 45.64%), mixed (n=28, 18.79%), and cholestatic (n=53, 35.57%). The proportion of splenomegaly in patients with cholestatic injury (56.60%) was significantly higher than that in those with hepatocellular (35.71%) and mixed injury (22.06%) (PConclusions: Statistical differences are observed in the serum biomarkers of DILI according to varying biochemical damage and degree of severity. Q-LSC and ascites were associated with the severity of DILI, and a combined model incorporating Q-LSC and ascites can effectively predict high-risk DILI.

  • Open Access

    Uterine arteriovenous fistula: A case report and literature review

    by Siyu Chen, Xia Song, Jinzhan Su, Fang Wu, Dongying Su, Shufeng Fan, Hui Hu, Jie Li, Miaoer Li, Jianmin Bs Wen

    Rationale: Uterine arteriovenous fistula (UAVF) is not very common and can occur after traumatic activities, such as surgery, delivery, miscarriage, and curettage. Previous studies have reported on the ultrasound and digital subtraction angiography (DSA) findings of UAVF; however, there are few reports on computed tomography (CT), CT angiography (CTA), and magnetic resonance imaging (MRI) findings. We report a case in which residual products of conception were diagnosed during an MRI, with an oversight of the presence of UAVF. Patient concerns: A female aged 34 years presented with intermittent vaginal bleeding for 15 days that had increased in the last 2 hours, 43 days after an abortion. Ultrasound showed abnormal echo masses in the right uterine horn. MRI examination showed incomplete pregnancy. CT, CTA, and DSA examinations showed incomplete pregnancy with UAVF. Diagnoses: The postoperative combined pathological diagnosis was right uterine horn insufficiency abortion with UAVF. Interventions: To reduce intraoperative bleeding, interventional embolization was performed, followed by surgical resection. Outcomes: Recovery was good, and the patient was discharged. Lessons: After an induced abortion, especially after multiple abortions, in patients presenting with abdominal pain, irregular vaginal bleeding, sudden heavy bleeding, and other phenomena, we should consider the possibility of UAVF. In this paper, we report using CT, CTA, MRI, and DSA imaging findings of UAVF to improve our understanding of UAVF imaging performance. For patients with persistent bleeding and fertility requirements that require immediate surgical resection of the lesion, interventional embolization and resection of the lesion can be selected to reduce the risk of intraoperative bleeding and recurrence.


  • Open Access

    The value of Gd-EOB-DTPA-enhanced MRI in the differential diagnosis of HCC with hyperintensity on HBP and FNH

    by Xin-Hui Bs Zhuang, Dong-Ying Su, Miao-Er Li, Jinzhan Su, Shu-Feng Fan, Fang Wu

    The aim was to investigate the differential diagnostic potential of gadolinium-ethoxybenzyl-diethylenetriamine pentaacetic acid (Gd-EOB-DTPA)-enhanced magnetic resonance imaging (MRI) features and establish a nomogram model for distinguishing hepatocellular carcinoma (HCC) from focal nodular hyperplasia (FNH) presenting with hyperintensity on the hepatobiliary phase (HBP). This retrospective study enrolled 80 patients with pathologically confirmed HCC or FNH who underwent Gd-EOB-DTPA-enhanced MRI between January 2017 and December 2020. All lesions exhibited hyperintensity on HBP. Morphological characteristics, signal patterns, and apparent diffusion coefficient (ADC) values were analyzed. Univariate and multivariate logistic regression analyses were performed to identify independent predictors of HCC, adjusting for age and sex. A diagnostic nomogram was subsequently constructed. After adjusting for age and sex, the study showed that nodule-in-nodule hyperintensity on HPB (odds ratio [OR] = 36.46, 95% confidence interval [CI]: 4.01–331.13), an ADC value ≤1.087 × 10−3 mm2/s (OR = 0.004, 95% CI: 0.00–0.06), and the absence of a central scar (OR = 0.04, 95% CI: 0.003–0.40) were independent predictors of HCC. The nomogram incorporating these predictors demonstrated excellent diagnostic performance, with an area under the receiver operating characteristic curve of 0.933 (95% CI: 0.874–0.991). The calibration curve showed optimal agreement between predicted and observed probabilities. Gd-EOB-DTPA-enhanced MRI characteristics combined with ADC values enable reliable differentiation between HBP-hyperintense HCC and FNH. The proposed nomogram model provides a clinically applicable tool for improving diagnostic accuracy in challenging cases.


  • Open Access

    Quantitative Multi-Parameter MRI Evaluation of Hashimoto\'s Thyroiditis Changes After Dietary Interventions

    by Ning Dai, Qi-Hao Shi, Long-Wei Zheng, Xiao-Shan Huang, Shu-Feng Fan

    Quantitative Multi-Parameter MRI Evaluation of Hashimoto's Thyroiditis Changes After Dietary Interventions - Article abstract #947862