Filters








58 Hits in 1.1 sec

Encoding CT Anatomy Knowledge for Unpaired Chest X-ray Image Decomposition [article]

Zeju Li, Han Li, Hu Han, Gonglei Shi, Jiannan Wang, S. Kevin Zhou
2019 arXiv   pre-print
Although chest X-ray (CXR) offers a 2D projection with overlapped anatomies, it is widely used for clinical diagnosis. There is clinical evidence supporting that decomposing an X-ray image into different components (e.g., bone, lung and soft tissue) improves diagnostic value. We hereby propose a decomposition generative adversarial network (DecGAN) to anatomically decompose a CXR image but with unpaired data. We leverage the anatomy knowledge embedded in CT, which features a 3D volume with
more » ... ly visible anatomies. Our key idea is to embed CT priori decomposition knowledge into the latent space of unpaired CXR autoencoder. Specifically, we train DecGAN with a decomposition loss, adversarial losses, cycle-consistency losses and a mask loss to guarantee that the decomposed results of the latent space preserve realistic body structures. Extensive experiments demonstrate that DecGAN provides superior unsupervised CXR bone suppression results and the feasibility of modulating CXR components by latent space disentanglement. Furthermore, we illustrate the diagnostic value of DecGAN and demonstrate that it outperforms the state-of-the-art approaches in terms of predicting 11 out of 14 common lung diseases.
arXiv:1909.12922v1 fatcat:vmdcbwa7ebabtbfrbw427x2jeq

Event Encryption for Neuromorphic Vision Sensors: Framework, Algorithm, and Evaluation

Bowen Du, Weiqi Li, Zeju Wang, Manxin Xu, Tianchen Gao, Jiajie Li, Hongkai Wen
2021 Sensors  
Nowadays, our lives have benefited from various vision-based applications, such as video surveillance, human identification and aided driving. Unauthorized access to the vision-related data greatly threatens users' privacy, and many encryption schemes have been proposed to secure images and videos in those conventional scenarios. Neuromorphic vision sensor (NVS) is a brand new kind of bio-inspired sensor that can generate a stream of impulse-like events rather than synchronized image frames,
more » ... ch reduces the sensor's latency and broadens the applications in surveillance and identification. However, the privacy issue related to NVS remains a significant challenge. For example, some image reconstruction and human identification approaches may expose privacy-related information from NVS events. This work is the first to investigate the privacy of NVS. We firstly analyze the possible security attacks to NVS, including grayscale image reconstruction and privacy-related classification. We then propose a dedicated encryption framework for NVS, which incorporates a 2D chaotic mapping to scramble the positions of events and flip their polarities. In addition, an updating score has been designed for controlling the frequency of execution, which supports efficient encryption on different platforms. Finally, extensive experiments have demonstrated that the proposed encryption framework can effectively protect NVS events against grayscale image reconstruction and human identification, and meanwhile, achieve high efficiency on various platforms including resource-constrained devices.
doi:10.3390/s21134320 pmid:34202626 pmcid:PMC8272045 fatcat:eutwbh2tmnaztne7l4gauryytq

Deep Generative Adversarial Networks for Thinsection Infant MR Image Reconstruction

Jiaqi Gu, Zeju Li, Yuanyuan Wang, Haowei Yang, Zhongwei Qiao, Jinhua Yu
2019 IEEE Access  
The key point of detail recalibration lies in information reuse.  ... 
doi:10.1109/access.2019.2918926 fatcat:nbq22prw3bgc3m3xnebesbudfa

Overfitting of neural nets under class imbalance: Analysis and improvements for segmentation [article]

Zeju Li and Konstantinos Kamnitsas and Ben Glocker
2019 arXiv   pre-print
Overfitting in deep learning has been the focus of a number of recent works, yet its exact impact on the behavior of neural networks is not well understood. This study analyzes overfitting by examining how the distribution of logits alters in relation to how much the model overfits. Specifically, we find that when training with few data samples, the distribution of logit activations when processing unseen test samples of an under-represented class tends to shift towards and even across the
more » ... ion boundary, while the over-represented class seems unaffected. In image segmentation, foreground samples are often heavily under-represented. We observe that sensitivity of the model drops as a result of overfitting, while precision remains mostly stable. Based on our analysis, we derive asymmetric modifications of existing loss functions and regularizers including a large margin loss, focal loss, adversarial training and mixup, which specifically aim at reducing the shift observed when embedding unseen samples of the under-represented class. We study the case of binary segmentation of brain tumor core and show that our proposed simple modifications lead to significantly improved segmentation performance over the symmetric variants.
arXiv:1907.10982v2 fatcat:fo7ksdtdknfwlfcrksooymgmka

Biological Function and Application of Picornaviral 2B Protein: A New Target for Antiviral Drug Development

Zengbin Li, Zixiao Zou, Zeju Jiang, Xiaotian Huang, Qiong Liu
2019 Viruses  
Li et al.  ... 
doi:10.3390/v11060510 pmid:31167361 pmcid:PMC6630369 fatcat:hfjxx3hvmvfctkwq4ppojjwtye

Deep Learning based Radiomics (DLR) and its usage in noninvasive IDH1 prediction for low grade glioma

Zeju Li, Yuanyuan Wang, Jinhua Yu, Yi Guo, Wei Cao
2017 Scientific Reports  
Author Contributions Zeju Li, Yuanyuan Wang, Jinhua Yu and Yi Guo conceived of the project, analyzed the data and wrote the paper.  ... 
doi:10.1038/s41598-017-05848-2 pmid:28710497 pmcid:PMC5511238 fatcat:hy4cbfi2avgbvft2skvgxekaaq

Low-Grade Glioma Segmentation Based on CNN with Fully Connected CRF

Zeju Li, Yuanyuan Wang, Jinhua Yu, Zhifeng Shi, Yi Guo, Liang Chen, Ying Mao
2017 Journal of Healthcare Engineering  
This work proposed a novel automatic three-dimensional (3D) magnetic resonance imaging (MRI) segmentation method which would be widely used in the clinical diagnosis of the most common and aggressive brain tumor, namely, glioma. The method combined a multipathway convolutional neural network (CNN) and fully connected conditional random field (CRF). Firstly, 3D information was introduced into the CNN which makes more accurate recognition of glioma with low contrast. Then, fully connected CRF was
more » ... added as a postprocessing step which purposed more delicate delineation of glioma boundary. The method was applied to T2flair MRI images of 160 low-grade glioma patients. With 59 cases of data training and manual segmentation as the ground truth, the Dice similarity coefficient (DSC) of our method was 0.85 for the test set of 101 MRI images. The results of our method were better than those of another state-of-the-art CNN method, which gained the DSC of 0.76 for the same dataset. It proved that our method could produce better results for the segmentation of low-grade gliomas.
doi:10.1155/2017/9283480 pmid:29065666 pmcid:PMC5485483 fatcat:74m4ouulrjcvhijs7ztnhtq3b4

Efficacy and Safety of Oncolytic Viruses in Randomized Controlled Trials: A Systematic Review and Meta-Analysis

Zengbin Li, Zeju Jiang, Yingxuan Zhang, Xiaotian Huang, Qiong Liu
2020 Cancers  
Oncolytic virotherapy is a promising antitumor therapeutic strategy. It is based on the ability of viruses to selectively kill cancer cells and induce host antitumor immune responses. However, the clinical outcomes of oncolytic viruses (OVs) vary widely. Therefore, we performed a meta-analysis to illustrate the efficacy and safety of oncolytic viruses. The Cochrane Library, PubMed, and EMBASE databases were searched for randomized controlled trials (RCTs) published up to January 31, 2020. The
more » ... ta for objective response rate (ORR), overall survival (OS), progression-free survival (PFS), and adverse events (AEs) were independently extracted by two investigators from 11 studies that met the inclusion criteria. In subgroup analyses, the objective response rate benefit was observed in patients treated with oncolytic DNA viruses (odds ratio (OR) = 4.05; 95% confidence interval (CI): 1.96–8.33; p = 0.0002), but not in those treated with oncolytic RNA viruses (OR = 1.00, 95% CI: 0.66–1.52, p = 0.99). Moreover, the intratumoral injection arm yielded a statistically significant improvement (OR = 4.05, 95% CI: 1.96–8.33, p = 0.0002), but no such improvement was observed for the intravenous injection arm (OR = 1.00, 95% CI: 0.66–1.52, p = 0.99). Among the five OVs investigated in RCTs, only talimogene laherparepvec (T-VEC) effectively prolonged the OS of patients (hazard ratio (HR), 0.79; 95% CI: 0.63–0.99; p = 0.04). None of the oncolytic virotherapies improved the PFS (HR = 1.00, 95% CI: 0.85–1.19, p = 0.96). Notably, the pooled rate of severe AEs (grade ≥3) was higher for the oncolytic virotherapy group (39%) compared with the control group (27%) (risk difference (RD), 12%; risk ratio (RR), 1.44; 95% CI: 1.17–1.78; p = 0.0006). This review offers a reference for fundamental research and clinical treatment of oncolytic viruses. Further randomized controlled trials are needed to verify these results.
doi:10.3390/cancers12061416 pmid:32486227 pmcid:PMC7352817 fatcat:63v3olxpafdg5kkalv7x5nrsi4

Enhancing MR Image Segmentation with Realistic Adversarial Data Augmentation [article]

Chen Chen, Chen Qin, Cheng Ouyang, Zeju Li, Shuo Wang, Huaqi Qiu, Liang Chen, Giacomo Tarroni, Wenjia Bai, Daniel Rueckert
2022 arXiv   pre-print
., 2019) , brain segmentation (Li et al., 2019) , and cardiac segmentation (Chaitanya et al., 2019) .  ...  This is a common issue in consistency-regularized methods Li et al. (2020); Berthelot et al. (2019).  ... 
arXiv:2108.03429v2 fatcat:m24wykdkbna3fdtq2t5qdlgq2i

Safety of catheter-directed thrombolysis for the treatment of acute lower extremity deep vein thrombosis

Li Wang, Chuanlin Zhang, Shaoyu Mu, Chao Hsing Yeh, Liqun Chen, Zeju Zhang, Xueqin Wang
2017 Medicine  
Despite established guidelines, catheter-directed thrombolysis (CDT) for the management of acute lower extremity deep vein thrombosis (DVT) should not be overstated because the risks of CDT are uncertain. We performed a meta-analysis to comprehensively and quantitatively evaluate the safety of CDT for patients with acute lower extremity DVT. Methods: Relevant databases, including PubMed, Embase, Cochrane, Ovid MEDLINE, and Scopus, were searched up to January 2017. The inclusion criteria were
more » ... lied to select patients with acute lower extremity DVT treated by CDT or compared CDT with anticoagulation. In case series studies, the pooled estimates of safety outcomes for complications, pulmonary embolism (PE), and mortality were calculated across studies. In studies comparing CDT with anticoagulation, summary odds ratios (ORs) were calculated. Results: Of the 1696 citations identified, 24 studies (6 comparing CDT with anticoagulation and 18 case series) including 9157 patients met the eligibility criteria. In the case series studies, the pooled risks of major, minor, and total complications were 0.03 (95% confidence interval [CI]: 0.02-0.04), 0.07 (95% CI: 0.05-0.08), and 0.09 (95% CI: 0.08-0.11), respectively; other pooled risk results were 0.00 for PE (95% CI: 0.00-0.01) and 0.07 for mortality (95% CI: 0.03-0.11). Our meta-analysis of 6 studies comparing the risk of complications and PE related to CDT with those related to anticoagulation showed that CDT was associated with an increased risk of complications (OR = 4.36; 95% CI: 2.94-6.47) and PE (OR = 1.57; 95% CI: 1.37-1.79). Conclusion: Acute lower extremity DVT patients receiving CDT are associated with a low risk of complications. However, compared with anticoagulation, CDT is associated with a higher risk of complications and PE. Rare mortality related to thrombolytic therapy was reported. More evidence should be accumulated to prove the safety of CDT. Abbreviations: CDT = catheter-directed thrombolysis, CI = confidence interval, DVT = deep vein thrombosis, OR = odds ratio, PE = pulmonary embolism, RCT = randomized clinical trial, VTE = venous thromboembolism.
doi:10.1097/md.0000000000007922 pmid:28858115 pmcid:PMC5585509 fatcat:oa7btnsxqrbetaowgzdcnvftai

Left Ventricle Segmentation via Optical-Flow-Net from Short-axis Cine MRI: Preserving the Temporal Coherence of Cardiac Motion [article]

Wenjun Yan, Yuanyuan Wang, Zeju Li, Rob J. van der Geest, Qian Tao
2018 arXiv   pre-print
Quantitative assessment of left ventricle (LV) function from cine MRI has significant diagnostic and prognostic value for cardiovascular disease patients. The temporal movement of LV provides essential information on the contracting/relaxing pattern of heart, which is keenly evaluated by clinical experts in clinical practice. Inspired by the expert way of viewing Cine MRI, we propose a new CNN module that is able to incorporate the temporal information into LV segmentation from cine MRI. In the
more » ... proposed CNN, the optical flow (OF) between neighboring frames is integrated and aggregated at feature level, such that temporal coherence in cardiac motion can be taken into account during segmentation. The proposed module is integrated into the U-net architecture without need of additional training. Furthermore, dilated convolution is introduced to improve the spatial accuracy of segmentation. Trained and tested on the Cardiac Atlas database, the proposed network resulted in a Dice index of 95% and an average perpendicular distance of 0.9 pixels for the middle LV contour, significantly outperforming the original U-net that processes each frame individually. Notably, the proposed method improved the temporal coherence of LV segmentation results, especially at the LV apex and base where the cardiac motion is difficult to follow.
arXiv:1810.08753v1 fatcat:5dtkrosuzrbd7dz7mt6vj3ma2m

Catheter-directed thrombolysis for patients with acute lower extremity deep vein thrombosis: a meta-analysis

Wang Li, Zhang Chuanlin, Mu Shaoyu, Chao Hsing Yeh, Chen Liqun, Zhang Zeju
2018 Revista Latino-Americana de Enfermagem  
Objectives: To evaluate case series studies that quantitatively assess the effects of catheter-directed thrombolysis (CDT) and compare the efficacy of CDT and anticoagulation in patients with acute lower extremity deep vein thrombosis (DVT). Methods: Relevant databases, including PubMed, Embase, Cochrane, Ovid MEDLINE and Scopus, were searched through January 2017. The inclusion criteria were applied to select patients with acute lower extremity DVT treated with CDT or with anticoagulation. In
more » ... he case series studies, the pooled estimates of efficacy outcomes for patency rate, complete lysis, rethrombosis and post-thrombotic syndrome (PTS) were calculated across the studies. In studies comparing CDT with anticoagulation, summary odds ratios (ORs) were calculated. Results: Twenty-five articles (six comparing CDT with anticoagulation and 19 case series) including 2254 patients met the eligibility criteria. In the case series studies, the pooled results were a patency rate of 0.87 (95% CI: 0.85-0.89), complete lysis 0.58 (95% CI: 0.40-0.75), rethrombosis 0.11 (95% CI: 0.06-0.17) and PTS 0.10 (95% CI: 0.08-0.12). Six studies comparing the efficacy outcomes of CDT and anticoagulation showed that CDT was associated with a reduction of PTS (OR 0.38, 95%CI 0.26-0.55, p<0.0001) and a higher patency rate (OR 4.76, 95%CI 2.14-10.56, p<0.0001). Conclusion: Acute lower extremity DVT patients receiving CDT were found to have a lower incidence of PTS and a higher incidence of patency rate. In our meta-analysis, CDT is shown to be an effective treatment for acute lower extremity DVT patients.
doi:10.1590/1518-8345.2309.2990 pmid:29947719 pmcid:PMC6047892 fatcat:zmygux3jzndozhfb2b7363hjpq

Cross-layer transmission and energy scheduling under full-duplex energy harvesting wireless OFDM joint transmission

Hongjia Li, Zejue Wang, Dan Hu, Song Ci, Zhen Xu
2016 Science China Information Sciences  
optimal algorithm and the online algorithm for brevity in the following, respectively.Downloaded to IP: 192.168.0.213 On: 2020-05-11 05:48:03 http://engine.scichina.com/doi/10.1007/s11432-015-5481-9 Li  ...  192.168.0.213 On: 2020-05-11 05:48:03 http://engine.scichina.com/doi/10.1007/s11432-015-5481-9 ) Downloaded to IP: 192.168.0.213 On: 2020-05-11 05:48:03 http://engine.scichina.com/doi/10.1007/s11432-015-5481-9 Li  ... 
doi:10.1007/s11432-015-5481-9 fatcat:hnocxkqsa5dd7ljymjhjfohywm

Causality-inspired Single-source Domain Generalization for Medical Image Segmentation [article]

Cheng Ouyang, Chen Chen, Surui Li, Zeju Li, Chen Qin, Wenjia Bai, Daniel Rueckert
2021 arXiv   pre-print
, Zeju Li, Wenjia Bai and Daniel Rueckert are with the Department of Computing, Imperial College London, analysis of acquisition shift.  ...  , Zeju Li, Chen Qin, Wenjia Bai and Daniel Rueckert Abstract—Deep learning models usually suffer from domain or device manufacturers) [10  ... 
arXiv:2111.12525v4 fatcat:e4ht2fhmijherf6jxo3fgmbgkm

Improved Long Ureteral Reconstruction With Ileum By Longitudinal Clipping And Mucosal Stripping: An Animal Study

Hao Gu, Shulian Chen, Yubo Wu, Lei Shen, Yihang Luo, Xu Li, Houjin Huang, Zeju Zhao
2020 Urology Journal  
To investigate the possibility of bridging long ureteral defects by longitudinal clipping and mucosal stripping of the pedicled segment of ileum (CMSPI). Ten beagle dogs (five males and females aged 2-3 years) were used to model a defect of the entire ureter. An ileal segment was selected, and half of the intestinal segment was longitudinally resected, without mesenteric resection. The intestinal mucosa was removed. Then, the ileum was sutured to form a tube connecting the renal pelvis to the
more » ... adder. A 5F ureter stent was inserted into the ileum and removed 4 weeks after surgery. Intravenous urography (IVU) was used to observe the reconstructed ureters at 6 and 12 weeks after the operation. Blood samples were collected before surgery and during each radiological examination to assess electrolyte and renal function. Five dogs were randomly euthanized after each IVU. After macroscopic analysis, hematoxylin-eosin (H&E) staining was performed to observe the microscopic changes in the reconstructed ureter. All dogs were in good condition after surgery. Changes in blood electrolyte and renal function after surgery were not significant (Cl- P = .595; Ur P = .852). IVU demonstrated no ureteral obstruction or extravasation of the contrast agent; however, mild hydronephrosis were observed in three dogs. Macroscopic analysis indicated that the reconstructed ureter was intact without strictures. H&E showed that no mucosal structure was present on the luminal surface. CMSPI is feasible for bridging long ureteral defects and has shown good efficacy in this preliminary study.
doi:10.22037/uj.v0i0.5330 pmid:31912471 fatcat:rabxjfjcqfe3bo5mxj7vj66wma
« Previous Showing results 1 — 15 out of 58 results