RHRenjiu Hu

Renjiu Hu

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

Current Organizations / Institutes

The organizations/institutes from the scholar's latest article


Renjiu Hu has a total of 86 co-authors. Here are the top co-authors by number of shared publications:


Latest Publications

  • Neural Optimization for Image Registration via Joint Modeling of Global Affine and Local Deformation Transformations

    by Xiang Chen, Renjiu Hu, Jiacheng Wang, Min Liu, Yaonan Wang, Jiazheng Wang, Rongguang Wang, Gaolei Li, Hang Zhang

    Conventional registration approaches frequently underperform when applied to sparse feature alignment (e.g., retinal vessels and filamentous collagen fibers in second-harmonic generation (SHG) and bright-field (BF) images), as these tasks demand simultaneous handling of global affine registration and local deformation correction. End-to-end learning-based approaches struggle with minimal effective gradients from loss back-propagation of these sparse features, while descriptor matching methods, though helpful, lack fidelity loss and fail to adapt to local deformation. To address these issues, we propose Neural Affine Optimization (NeOn), which implicitly approximates discrete optimization using a few neural network layers, combined with a sampling-regression layer to handle affine transformations. NeOn allows iterative refinement with fidelity loss and provides a flexible transition between a purely affine configuration and a linear weighted blend of affine and deformation fields. NeOn’s performance was validated on four public datasets. In multi-modal SHG-BF microscopy registration, NeOn achieved top rankings on the validation leaderboard for Task 3 of the Learn2Reg Challenge 2024. For retinal image registration, NeOn outperformed existing methods on both mono-modal and multi-modal datasets, reducing target registration error from 6.3 to 2.1 pixels in mono-modal and from 2.6 to 1.8 pixels in multi-modal registration. Furthermore, NeOn demonstrates strong generalization and can be effectively extended to 3D multi-modality image registration scenarios.

  • Open Access

    Improving Glioblastoma Classification Using Quantitative Transport Mapping with a Synthetic Data Trained Deep Neural Network

    by Dominick J Romano, Alexandra Grace Roberts, Benjamin Weppner, Qihao Zhang, Maneesh John, Renjiu Hu, Mert Şişman, Ilhami Kovanlikaya, Gloria C Chiang, Pascal Spincemaille, et al.

    Purpose: To develop a deep neural network-based, AIF-free, perfusion estimation method (QTMnet) for improved performance on glioma classification. Methods: A globally defined arterial input function (AIF) is needed to recover perfusion parameters in the two-compartment exchange model (2CXM). We have developed Quantitative Transport Mapping (QTM) to create an AIF-independent estimation method. QTM estimation can be formulated using deep neural networks trained on synthetic DCE-MRI data (QTMnet). Here, we provide a fluid mechanics-based DCE-MRI simulation with exchange between the capillaries and extravascular extracellular space. We implemented tumor ROI generation to morphologically characterize tissue perfusion. We compared our QTMnet implementation with 2CXM on 30 glioma human subjects, 15 of which had low-grade gliomas, and 15 with high-grade glioblastomas. Results: QTMnet outperforms (best AUC: 0.973) traditional 2CXM (best AUC: 0.911) in a glioma grading task. Conclusion: The AIF-independent QTMnet estimation provides a quantitative delineation between low-grade and high-grade gliomas.

  • Encoder-Only Image Registration

    by Xiang Chen, Renjiu Hu, Jinwei Zhang, Yuxi Zhang, Xinyao Yu, Min Liu, Yaonan Wang, Hang Zhang

    Learning-based techniques have significantly improved the accuracy and speed of deformable image registration. However, challenges such as reducing computational complexity and handling large deformations persist. To address these challenges, we analyze how convolutional neural networks (ConvNets) influence registration performance using the Horn-Schunck optical flow equation. Supported by prior studies and our empirical experiments, we observe that ConvNets play two key roles in registration: linearizing local intensities and harmonizing global contrast variations. Guided by these insights, we propose the Encoder-Only Image Registration (EOIR) framework comprising five modifications to existing approaches, to achieve a better accuracy-efficiency trade-off. EOIR separates feature learning from flow estimation, employing only a 3-layer ConvNet for feature extraction and a set of 3-layer flow estimators to construct a Laplacian feature pyramid, progressively composing diffeomorphic deformations under a large-deformation model. Results on six datasets across different modalities and anatomical regions demonstrate EOIR’s effectiveness, achieving superior accuracy-efficiency and accuracy-smoothness trade-offs. With comparable accuracy, EOIR provides better efficiency and smoothness, and vice versa. The source code of EOIR is available on Github.


  • Validation of quantitative transport mapping (QTM) with an ex vivo perfused liver model

    by Dominick Romano, Qihao Zhang, Alexandra G. Roberts, Benjamin Weppner, Jian Shou, Ilhami Kovanlikaya, Mert Şişman, Renjiu Hu, Thanh D. Nguyen, Alexey V. Dimov, et al.

    Purpose: To evaluate the accuracy of a deep learning–based quantitative transport mapping method (QTMnet) for measuring total tissue perfusion. Methods: QTMnet obtains tissue perfusion parametric maps from dynamic contrast‐enhanced MRI images by training on simulated data. This data uses synthetic arterial and venous vasculature geometries with flow based on constrained constructive optimization. Gadolinium contrast agent distribution is governed by the transport‐forward problem, allowing us to generate a synthetic concentration spacetime profile for a given flow, blood volume fraction, and boundary condition. Tissue flows determined by QTMnet were compared to those obtained with traditional perfusion quantification (Kety equation with Tofts generalization), which uses a global arterial input function. Their total flow accuracies were validated on explanted porcine livers that were connected to an MR compatible flow pump with specified total flow rate for dynamic contrast‐enhanced MRI experiments. Results: The mean total flow error for QTMnet was −0.34% ± 16.21% with range [−24.79%, 23.96%], compared to −35.74% ± 36.30% [−77.28%, 29.87%] for the Kety method. QTMnet provides 72% lower mean absolute error than the Kety method (12.15% vs. 43.21%, a 3.6‐fold reduction). Conclusion: The fluid mechanics‐based QTMnet accurately estimates total tissue flow in liver explants.


  • High Through-Thickness Thermal Conductivity in an Edge-On Two-Dimensional Polyamide Thin Film

    by Xiuqiang Li, Guangxin Lv, Yinglong Hu, Yu-Hsuan Tsao, Renjiu Hu, Zhiting Tian, Kejun Liu, Hao Ma

    High thermal conductivity is essential for polymer applications such as electronic chip encapsulation, where efficient heat dissipation ensures system functionality and reliability. Here, we introduce a novel strategy to enhance through-plane thermal conductivity in 2D covalent organic frameworks (COFs). A highly crystalline edge-on 2D polyamide (v2DPA) film achieves a thermal conductivity of 1.16 ± 0.05 W/(mK) at 310 K, surpassing the previous record (1.03 W/(mK) in COF-5 [Evans et al. Nat. Mater. 2021, 20, 1142]) and aligning with molecular dynamics predictions (1.11 ± 0.07 W/(mK)). This value is nearly three times higher than that of bulk PA (0.34 ± 0.03 W/(mK)). Phonon dispersion calculations attribute this enhancement to strong covalent bonding, increasing phonon lifetimes, and group velocities. Our findings highlight the effectiveness of orienting 2D polymer and layer-stacked 2D COF films in an edge-on configuration to improve through-thickness thermal conductivity, offering a promising pathway for their integration into electronic thermal management applications.


  • Spatially Covariant Image Registration With Text Prompts

    by Xiang Chen, Min Liu, Rongguang Wang, Renjiu Hu, Dongdong Liu, Gaolei Li, Yaonan Wang, Hang Zhang

    Medical images are often characterized by their structured anatomical representations and spatially inhomogeneous contrasts. Leveraging anatomical priors in neural networks can greatly enhance their utility in resource-constrained clinical settings. Prior research has harnessed such information for image segmentation, yet progress in deformable image registration has been modest. Our work introduces textSCF, a novel method that integrates spatially covariant filters and textual anatomical prompts encoded by visual-language models, to fill this gap. This approach optimizes an implicit function that correlates text embeddings of anatomical regions to filter weights. textSCF not only boosts computational efficiency but can also retain or improve registration accuracy. By capturing the contextual interplay between anatomical regions, it offers impressive interregional transferability and the ability to preserve structural discontinuities during registration. textSCF’s performance has been rigorously tested on intersubject brain magnetic resonance imaging (MRI) and abdominal computerized tomography (CT) registration tasks, outperforming existing state-of-the-art models in the MICCAI Learn2Reg 2021 challenge and leading the leaderboard. In abdominal registrations, textSCF’s larger model variant improved the Dice score by 11.3% over the second-best model, while its smaller variant maintained similar accuracy but with an 89.13% reduction in network parameters and a 98.34% decrease in computational operations.


  • Open Access

    Large Scale Unsupervised Brain MRI Image Registration Solution for Learn2Reg 2024

    by Yuxi Zhang, Xiang Chen, Jiazheng Wang, Min Liu, Yaonan Wang, Dongdong Liu, Renjiu Hu, Hang Zhang

    In this paper, we summarize the methods and experimental results we proposed for Task 2 in the learn2reg 2024 Challenge. This task focuses on unsupervised registration of anatomical structures in brain MRI images between different patients. The difficulty lies in: (1) without segmentation labels, and (2) a large amount of data. To address these challenges, we built an efficient backbone network and explored several schemes to further enhance registration accuracy. Under the guidance of the NCC loss function and smoothness regularization loss function, we obtained a smooth and reasonable deformation field. According to the leaderboard, our method achieved a Dice coefficient of 77.34%, which is 1.4% higher than the TransMorph. Overall, we won second place on the leaderboard for Task 2.

  • Open Access

    Anomalous lattice thermal conductivity increase with temperature in cubic GeTe correlated with strengthening of second-nearest neighbor bonds

    by Samuel Kielar, Chen Li, Han Huang, Renjiu Hu, Carla Slebodnick, Ahmet Alatas, Zhiting Tian

    Understanding thermal transport mechanisms in phase change materials is critical to elucidating the microscopic picture of phase transitions and advancing thermal energy conversion and storage. Experiments consistently show that cubic phase germanium telluride (GeTe) has an unexpected increase in lattice thermal conductivity with rising temperature. Despite its ubiquity, resolving its origin has remained elusive. In this work, we carry out temperature-dependent lattice thermal conductivity calculations for cubic GeTe through efficient, high-order machine-learned models and additional corrections for coherence effects. We corroborate the calculated phonon properties with our inelastic X-ray scattering measurements. Our calculated lattice thermal conductivity values agree well with experiments and show a similar increasing trend. Through additional bonding strength calculations, we propose that a major contributor to the increasing lattice thermal conductivity is the strengthening of second-nearest neighbor interactions. The findings herein serve to deepen our understanding of thermal transport in phase-change materials.


  • Rapid Photothermal Healing of Vitrimer Nanocomposites Activated by Gold-Nanoparticle-Coated Graphene Nanoplatelets

    by Yixin Ren, Amber M. Hubbard, Drake Austin, Jinghang Dai, Chen Li, Renjiu Hu, Peter Papaioannou, Catalin R. Picu, Dominik Konkolewicz, Alireza Sarvestani, et al.