Dong Yu

Dong Yu

Auto-generated profile, based on publicly available information

Dong Yu

Institute of Information Science, School of Computer Science and Technology, Beijing Jiaotong University, Beijing, China

Research Focus & Metrics

Academic Career Length

6Years (2021-Present)

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

Representative Publications

View All Publications
8 April 2021 in Ear & Hearing
Verified
Journal Article
Open access
YDYu Dong
Yu Dong
JBJeroen J. Briaire
Jeroen J. Briaire
MSMichael Siebrecht
Michael Siebrecht
HSH. Christiaan Stronks
H. Christiaan Stronks
JFJohan HM Frijns
Johan HM Frijns
Objectives: Misplacement of the electrode array is associated with impaired speech perception in patients with cochlear implants (CIs). Translocation of the electrode array is the most common misplacement. When a CI is translocated, it crosses the basilar membrane from the scala tympani into the scala vestibuli. The position of the implant can be determined on a postoperative CT scan. However, such a scan is not obtained routinely after CI insertion in many hospitals, due to radiation exposure and processing time. Previous studies have shown that impedance measures might provide information on the placement of the electrode arrays. The electrode impedance was measured by dividing the plateau voltage at the end of the first phase of the pulse by the injected current. The access resistance was calculated using the so-called access voltage at the first sampled time point after the start of the pulse divided by the injected current. In our study, we obtained the electrode impedance and the access resistance to detect electrode translocations using electrical field imaging. We have investigated how reliably these two measurements can detect electrode translocation, and which method performed best. Design: We calculated the electrode impedances and access resistances using electrical field imaging recordings from 100 HiFocus Mid-Scala CI (Advanced Bionics, Sylmar, CA) recipients. We estimated the normal values of these two measurements as the baselines of the implant placed in the cochlea without translocation. Next, we calculated the maximal electrode impedance deviation and the maximal access-resistance deviation from the respective baselines as predictors of translocation. We classified these two predictors as translocations or nontranslocations based on the bootstrap sampling method and receiver operating characteristics curves analysis. The accuracy could be calculated by comparing those predictive results to a gold standard, namely the clinical CT scans. To determine which measurement more accurately detected translocation, the difference between the accuracies of the two measurements was calculated. Results: Using the bootstrap sampling method and receiver operating characteristics–based optimized threshold criteria, the 95% confidence intervals of the accuracies of translocation detections ranged from 77.8% to 82.1% and from 89.5% to 91.2% for the electrode impedance and access resistance, respectively. The accuracies of the maximal access-resistance deviations were significantly larger than that of the maximal electrode impedance deviations. The location of the translocation as predicted by the access resistance was significantly correlated with the result derived from the CT scans. In contrast, no significant correlation was observed for the electrode impedance. Conclusions: Both the electrode impedance and access resistance proved reliable metrics to detect translocations for HiFocus Mid-Scala electrode arrays. The access resistance had, however, significantly better accuracy and it also reliably detected the electrode-location of translocations. The electrode impedance did not correlate significantly with the location of translocation. Measuring the access resistance is, therefore, the recommended method to detect electrode-array translocations. These measures can provide prompt feedback for surgeons after insertion, improving their surgical skills, and ultimately reducing the number of translocations. In the future, such measurements may allow near-real-time monitoring of the electrode array during insertion, helping to avoid translocations.

12 January 2023 in IEEE Access
Verified
Journal Article
Open access
DYDong Yu
Dong Yu
SLSuzhen Lin
Suzhen Lin
XLXiaofei Lu
Xiaofei Lu
DLDawei Li
Dawei Li
YWYanbo Wang
Yanbo Wang
This study proposes a new deep neural network to solve the multi-band image synchronous fusion problem (MBF-Net). Unlike other deep learning-based methods, our network architecture design combines the ideas of model-driven and data-driven methods, so it is more interpretable. First, a new multi-band image synchronous fusion model is proposed. The source image in the data fidelity terms and the prior regularization are implicitly represented by the deep learning network and jointly learned from the training data. The proposed model is then solved using a half quadratic splitting (HQS) algorithm and unfolded into a deep fusion network. In addition, a new saliency loss function is proposed to retain thermal radiation information to enhance the fusion effect. Finally, the experimental results on the TNO dataset demonstrated the effectiveness of the proposed MBF-Net.

4 February 2026 in IEEE Transactions on Multimedia
Verified
Journal Article
Closed access
DYDong Yu
Dong Yu
YTYepeng Tang
Yepeng Tang
CZChunjie Zhang
Chunjie Zhang
WWWei Wang
Wei Wang
Guodong YangGuodong Yang
Guodong Yang
YZYao Zhao
Yao Zhao
Image fusion aims to integrate multi-modal data, enhancing information completeness while reducing redundancy. However, image fusion faces the challenge of information imbalance. This imbalance is reflected in both intra-image variations, where different regions exhibit significant differences in information density and importance, and cross-modal inconsistencies, where corresponding regions in different modalities contribute unequally to the fused result. Most existing image fusion methods, such as models based on CNN and window attention mechanism, adopt a fixed approach that collects the same amount of neighboring information for each pixel. However, this fixed approach lacks adaptability to local variations in information density, ultimately limiting fusion performance. To this end, we propose IA2GNN, an imbalance-aware adaptive graph neural network designed for image fusion. To tackle intra- and cross-modal information imbalance, we leverage the flexibility of graph structures to dynamically adjust node connections during feature extraction and fusion. Specifically, we adjust the connectivity of nodes in the graph to allow regions with rich details to establish more connections, thereby enhancing the extraction and preservation of key information in these regions. For less informative regions, we reduce connections to prevent over-modeling. In this way, we optimize information distribution, achieving a balance between detail retention and information integration in the fusion results. Experimental results demonstrate that the proposed model outperforms baseline methods across multiple image fusion tasks, achieving up to +0.058 improvement in SSIM and +0.351 gain in MI.

1 July 2022
Verified
Conference Paper
Open access
YDYu Dong
Yu Dong
YLYue Li
Yue Li
DCDongqi Cui
Dongqi Cui
Jianbo GaoJianbo Gao
Jianbo Gao
ZGZhi Guan
Zhi Guan
ZCZhong Chen
Zhong Chen

17 June 2026 in Machine Intelligence Research
Verified
Journal Article
Closed access
DYDong Yu
Dong Yu
CZChunjie Zhang
Chunjie Zhang
XZXiao-Yu Zhang
Xiao-Yu Zhang
GZGaopeng Zhang
Gaopeng Zhang
YZYao Zhao
Yao Zhao

16 March 2023 in Entertainment Computing
Verified
Journal Article
Closed access
YDYu Dong
Yu Dong