Zhou Liu

Dr. Zhou Liu


Research Focus & Metrics

Academic Career Length

4Years (2023-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
11 August 2023 in European Radiology
Verified
Journal Article
Closed access
PDF file
HLHonghong Luo
Honghong Luo
LZLiyan Zou
Liyan Zou
QYQian Yang
Qian Yang
CYCuiyun Yuan
Cuiyun Yuan
KMKun Ma
Kun Ma
ZLZhou Liu
Zhou Liu

30 May 2023 in Pharmaceutics
Verified
Journal Article
Open access
PDF file
LLLing Liu
Ling Liu
SYShangpo Yang
Shangpo Yang
ZZZiliang Zheng
Ziliang Zheng
QLQingshuang Li
Qingshuang Li
CLChen chen Liu
Chen chen Liu
DGDuyang Gao
Duyang Gao
The accurate diagnosis and treatment of prostate cancer at an early stage is crucial to reduce mortality rates. However, the limited availability of theranostic agents with active tumor-targeting abilities hinders imaging sensitivity and therapeutic efficiency. To address this challenge, we have developed biomimetic cell membrane-modified Fe2O3 nanoclusters implanted in polypyrrole (CM-LFPP), achieving photoacoustic/magnetic resonance dual-modal imaging-guided photothermal therapy of prostate cancer. The CM-LFPP exhibits strong absorption in the second near-infrared window (NIR-II, 1000–1700 nm), showing high photothermal conversion efficiency of up to 78.7% under 1064 nm laser irradiation, excellent photoacoustic imaging capabilities, and good magnetic resonance imaging ability with a T2 relaxivity of up to 48.7 s−1 mM−1. Furthermore, the lipid encapsulation and biomimetic cell membrane modification enable CM-LFPP to actively target tumors, leading to a high signal-to-background ratio of ~30.2 for NIR-II photoacoustic imaging. Moreover, the biocompatible CM-LFPP enables low-dose (0.6 W cm−2) photothermal therapy of tumors under 1064 nm laser irradiation. This technology offers a promising theranostic agent with remarkable photothermal conversion efficiency in the NIR-II window, providing highly sensitive photoacoustic/magnetic resonance imaging-guided prostate cancer therapy.

1 September 2024 in Journal of Thoracic Disease
Verified
Journal Article
Open access
PDF file
LLLi Li
Li Li
QYQian Yang
Qian Yang
DLDehong Luo
Dehong Luo
XWXiaoliang Wang
Xiaoliang Wang
ZLZhou Liu
Zhou Liu
RHRong Huang
Rong Huang
Background: Preoperative accurate visceral pleural infiltration (VPI) diagnosis for T1-size non-small cell lung cancer (NSCLC) is significant for clinical decision-making. The study aimed to explore the diagnostic efficacy of computed tomography (CT) imaging features and serum biomarkers in diagnosing VPI in newly discovered subpleural NSCLC ≤3 cm. Methods: There were 447 patients with NSCLC ≤3 cm retrospectively enrolled and assigned to the VPI group (n=81) and the non-VPI group (n=366) based on elastic fiber staining results. The serum biomarkers and CT imaging features were obtained for each subject. Univariate and multivariate analyses were used to identify the independent predictors for VPI. Area under the receiver operating characteristic (ROC) curve (AUC) was used to evaluate the diagnostic performance of each independent predictor and combined predictors in predicting VPI, with performance compared using the DeLong test. Results: For tumor biomarkers, the VPI group had a significantly higher percentage of cases with abnormal carcino-embryonic antigen (CEA) level, cytokeratin 19 fragment (CYFRA21-1) level, and pro-gastrin-releasing peptide (ProGRP) level than that of the non-VPI group (P<0.001, P=0.003, P=0.004). However, in multivariate analysis, only the lesion-pleura relationship patterns type Ia [odds ratio (OR) =20.689; 95% confidence interval (CI): 5.058–84.622; P<0.001], type Ib (OR =5.155; 95% CI: 1.178–22.552; P=0.03), type II (OR =7.154; 95% CI: 1.733–29.53; P=0.007) with type III as reference, solid lesion density (OR =9.954; 95% CI: 4.976–19.911; P<0.001) with part-solid density as reference were identified as the independent predictors for VPI. In predicting VPI, the combined model (AUC =0.885) significantly outperformed models based on lesion density (AUC =0.833) and lesion-pleura relationship patterns (AUC =0.655) (all P<0.001). Conclusions: The CT predictors for VPI in patients with subpleural NSCLC (≤3 cm) were lesion density and lesion-pleura relationship patterns (pleural attachment and indentation), but not serum tumor biomarkers.

1 January 2024 in Quantitative Imaging in Medicine and Surgery
Verified
Journal Article
Open access
PDF file
RHRuibao Hu
Ruibao Hu
YXYongsheng Xie
Yongsheng Xie
LZLulu Zhang
Lulu Zhang
LLLijian Liu
Lijian Liu
HLHonghong Luo
Honghong Luo
ZHZhanli Hu
Zhanli Hu
Background: In low-dose computed tomography (LDCT) lung cancer screening, soft tissue is hardly appreciable due to high noise levels. While deep learning-based LDCT denoising methods have shown promise, they typically rely on structurally aligned synthesized paired data, which lack consideration of the clinical reality that there are no aligned LDCT and normal-dose CT (NDCT) images available. This study introduces an LDCT denoising method using clinically structure-unaligned but paired data sets (LDCT and NDCT scans from the same patients) to improve lesion detection during LDCT lung cancer screening. Methods: A cohort of 64 patients undergoing both LDCT and NDCT was randomly divided into training (n=46) and testing (n=18) sets. A two-stage training approach was adopted. First, Gaussian noise was added to NDCT data to create simulated LDCT data for generator training. Then, the model was trained on a clinically structure-unaligned paired data set using a Wasserstein generative adversarial network (WGAN) framework with the initial generator weights obtained during the first stage of training. An attention mechanism was also incorporated into the network. Results: Validated on a clinical CT data set, our proposed method outperformed other available methods [CycleGAN, Pixel2Pixel, block-matching and three-dimensional filtering (BM3D)] in noise removal and detail retention tasks in terms of the peak signal-to-noise ratio (PSNR), structural similarity index measure (SSIM), and root mean square error (RMSE) metrics. Compared with the results produced by BM3D, our method yielded an average improvement of approximately 7% in terms of the three evaluation indicators. The probability density profile of the denoised CT output produced using our method best fit the reference NDCT scan. Additionally, our two-stage model outperformed the one-stage WGAN-based model in both objective and subjective evaluations, further demonstrating the higher effectiveness of our two-stage training approach. Conclusions: The proposed method performed the best in removing noise from LDCT scans and exhibited good detail retention, which could potentially enhance the lesion detection and characterization effects obtained for soft tissues in the scanning scope of LDCT lung cancer screening.

27 May 2025 in IEEE Journal of Biomedical and Health Informatics
Verified
Journal Article
Closed access
HLHaizhou Liu
Haizhou Liu
YZYijia Zheng
Yijia Zheng
ZLZhou Liu
Zhou Liu
YJYuxi Jin
Yuxi Jin
ZLZhihua Li
Zhihua Li
ZHZhanli Hu
Zhanli Hu
Physiological and external motion cause inter-frame misalignment in chemical exchange saturation transfer magnetic resonance imaging (CEST-MRI), thereby compromising quantitative accuracy. In CEST-MRI, saturation effects induce intensity variations, resulting in motion-intensity coupling that makes registration particularly challenging. To address this issue, we extend the finite element digital image correlation (FE-DIC) framework by introducing an alternating correction strategy that iteratively refines both motion and intensity estimation. Unlike conventional FE-DIC approaches that assume intensity constancy, the proposed method incorporates mechanical regularization to suppress non-physical deformations, alongside intensity correction to compensate for reference-target contrast discrepancies. This mutual reinforcement enables progressively improved registration across the CEST sequence. The robustness and effectiveness of the method were evaluated on three datasets. In simulated liver data, it maintains RMSE within 0.4 pixels, reducing error by 0.5 pixels compared to RPCA&PCA (a PCA-based synthetic reference generation method for CEST registration). On clinical brain and pig cardiac data, it achieves average SSIM of 0.83, outperforming RPCA&PCA by 0.03 and surpassing CNN-based registration (e.g., AirLab) by 0.10. The consistent results across datasets highlight its generalizability, making it a promising tool for metabolic quantification in clinical and research settings.

1 July 2025 in Journal of Thoracic Disease
Verified
Journal Article
Open access
PDF file
JYJingwen Yu
Jingwen Yu
YZYihong Zhong
Yihong Zhong
YWYunfei Wang
Yunfei Wang
YZYan Zhang
Yan Zhang
XWXiaoliang Wang
Xiaoliang Wang
ZLZhou Liu
Zhou Liu
Background: Accurate preoperative diagnosis of pathological grades in T1-sized lung adenocarcinoma (LUAD) is crucial for clinical decision-making. The study aimed to investigate the value of dual-energy computed tomography (DECT) in distinguishing pathological grades in newly diagnosed LUAD lesions ≤3 cm in size. Methods: From October 2018 to January 2022, 137 patients with 161 pathologically confirmed LUAD lesions (≤3 cm) having received DECT were retrospectively enrolled with clinical information collected (low-grade: high-grade =41:120). CT values of monochromatic images at 40–140 keV, effective atomic number (Zeff), and iodine concentration of lesion (ICl) on plain, arterial phase (AP), and venous phase (VP) images were measured. Iodine concentration difference (ICD), normalized iodine concentration (NIC), and slope of the spectral curve (λHU) were further calculated. Difference tests and multiple stepwise regression were utilized to predict grades. Receiver operating characteristic (ROC) analysis was used to evaluate the diagnostic efficacy of the clinical model, spectral model, and combined model. Results: There were significant differences in CT values of 40–140 keV and Zeff between these two groups for the plain, AP, and VP images (all P0.05). Stepwise regression showed that CT40keV on plain spectral CT, CT140keV on AP, and CT140keV on VP are independently significant factors. The combined model achieved the best performance [area under the curve (AUC) =0.825], which significantly outperformed the clinical model (AUC =0.772, P=0.03), but not the spectral CT model (AUC =0.774, P=0.07). Conclusions: In addition to clinical features, single-energy CT values hold the potential to differentiate the pathological grade of T1-size LUAD.