ZLZhi Liu

Zhi Liu

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

Current Organizations / Institutes

The organizations/institutes from the scholar's latest article


Zhi Liu has a total of 31 co-authors. Here are the top co-authors by number of shared publications:


Latest Publications

  • Open Access

    Prognostic value of cardiac magnetic resonance pulmonary transit time and myocardial strain across different stages of ST-elevation myocardial infarction

    by Ying-Ying Hu, Zi-Xuan Zhang, Shuang Li, Guang-Yan Huo, Yan-Ling Wang, Zhi Liu, Nan Chen, Zheng Sun, Jie Lu

  • Open Access

    Cardiac CT Angiography in Diagnosing Patent Foramen Ovale: A Study on Patients with Suspected Patent Foramen Ovale-Associated Stroke

    by Lijie Sun, Chong Zheng, Zhenxing Fan, Jing Gao, Zhi Liu, Jin Si, Keling Xiao, Ming Yi, Haoyu Zhang, Jinghao Sun, et al.

    Background: Cardiac computed tomographic angiography (CTA) detects patent foramen ovale (PFO) with variable accuracy. This study investigated factors affecting CTA detectability for PFO in patients with suspected PFO-associated stroke. Methods: Consecutive patients with cryptogenic stroke and positive findings on contrast transcranial Doppler (cTCD) examinations were enrolled between November 2020 and April 2023 in this retrospective study. Each participant underwent transesophageal echocardiography (TEE) and cardiac CTA. Patients with confirmed PFO on TEE were categorized into two groups based on CTA detectability: the CTA-positive group (PFO identified by CTA) and the CTA-negative group (PFO missed by CTA). Univariate and multivariate logistic regression analyses were performed to identify predictors of CTA false-negative results. Results: Among 108 patients (mean age 46.7 ± 14.9 years, 47.2% male), the prevalence of PFO by TEE was 94.4% (102/108). Compared to TEE, cardiac CTA had a sensitivity of 70% (95% CI 61–79%), a specificity of 100% (95% CI 54–100%), a positive predictive value of 100% (95% CI 95–100%), and a negative predictive value of 16% (95% CI 6–32%). Among patients with PFO confirmed by TEE (n = 102), the incidence of moderate to large right-to-left shunts (RLS) was significantly higher in the CTA-positive group than in the CTA-negative (77.5% vs. 22.5%, p < 0.001). After adjusting for confounders, patients with moderate to large shunts showed a significantly lower likelihood of a CTA false-negative result compared to those with small shunts (OR 0.113, 95% CI 0.035–0.365, p < 0.001). In patients with moderate to large RLS, the sensitivity of cardiac CTA for diagnosing PFO increased to 90.16% (95% CI 82.69–97.64%). Conclusions: Cardiac CTA could be an effective complementary modality for selected patients with suspected PFO-associated stroke. Its diagnostic performance appears more reliable for identifying PFO in patients with moderate-to-large RLS than in those with small RLS.


  • Open Access

    Estimated glucose disposal rate (eGDR) predicts rehospitalization & death in acute decompensated heart failure in the elderly

    by Wenliang Zhai, Mingyi He, Chunyuan Wang, Jingyu He, Tao Wang, Xiurong Xing, Zhi Liu, Shubin Guo

    Objective This study aimed to explore the short-term predictive value of the estimated glucose disposal rate (eGDR) for rehospitalization or death in elderly patients with acute decompensated heart failure (ADHF). Methods 117 elderly patients with ADHF admitted to our hospital from January 2020 to January 2022 were recruited and divided into control (80 cases) and event (37 cases) groups based on whether they were rehospitalized or died within 90 days following discharge. Basic information, clinical characteristics, and laboratory test results of them were collected, and the eGDR was calculated. Multivariate logistic regression analysis identified the independent factors associated with the short-term prognosis, and receiver operating characteristic (ROC) curves assessed the predictive value of eGDR. Results The event group had significantly higher rates of comorbid diabetes (DM), hypertension, Class IV heart failure, and cardiogenic shock, as well as a larger waist circumference, higher peak troponin I during hospitalization, greater glycated hemoglobin (HbA1c) and fasting blood glucose (FBG), and lower eGDR values compared to the control group (P < 0.05). Multivariate logistic regression showed that comorbid hypertension, DM, Class IV cardiac function, cardiogenic shock, FBG, HbA1c, and eGDR were independent factors associated with poor short-term prognosis in elderly patients with ADHF (P < 0.05). ROC curve analysis showed that the area under the curve for eGDR in predicting poor short-term prognosis was 0.84 (P = 0.04), with an optimal cutoff value of 5.31, a sensitivity of 83.8%, and a specificity of 70.1%. Patients with eGDR < 5.31 had a significantly higher incidence of rehospitalization or death within 90 days compared to those with higher eGDR (52.0% vs. 16.5%, P < 0.001). Conclusion eGDR is an independent predictor of poor short-term prognosis in elderly patients with ADHF: a low eGDR was associated with a higher risk of rehospitalization and death within 90 days.

  • A Motion-Aware DNN Model with Edge Focus Loss and Quality Control for Short-Axis Left Ventricle Segmentation of Cine MR Sequences

    by Yu Wang, Zheng Sun, Zhi Liu, Jie Lu, Nan Zhang

    Accurate segmentation of the left ventricle myocardium is the key step of automatic assessment of cardiac function. However, the current methods mainly focus on the end-diastolic and the end-systolic frames in cine MR sequences and lack the attention to myocardial motion in the cardiac cycle. Additionally, due to the lack of fine segmentation tools, the simplified approach, excluding papillary muscles and trabeculae from myocardium, is applied in clinical practice. To solve these problems, we propose a motion-aware DNN model with edge focus loss and quality control in this paper. Specifically, the bidirectional ConvLSTM layer and a new motion attention layer are proposed to encode motion-aware feature maps, and an edge focus loss function is proposed to train the model to generate the fine segmentation results. Additionally, a quality control method is proposed to filter out the abnormal segmentations before subsequent analyses. Compared with state-of-the-art segmentation models on the public dataset and the in-house dataset, the proposed method has obtained high segmentation accuracy. On the 17-segment model, the proposed method has obtained the highest Pearson correlation coefficient at 14 of 17 segments, and the mean PCC of 85%. The experimental results highlight the segmentation accuracy of the proposed method as well as its availability to substitute for the manually annotated boundaries for the automatic assessment of cardiac function.