Zhenzhong Liu
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Previous Organizations / Institutes
The organizations / institutes from the scholar's other articles in the past ten years
Research Subjects
Research Interests
Zhenzhong Liu has a total of 81 co-authors. Here are the top co-authors by number of shared publications:
- Guobin Zhang
Tianjin Key Laboratory for Advanced Mechatronic System Design and Intelligent Control, Tianjin University of Technology, Tianjin, China
8 shared publications - Shuai Li
University of Oulu
5 shared publications - Runfeng Zhang
Tianjin Key Laboratory for Advanced Mechatronic System Design and Intelligent Control, School of Mechanical Engineering, Tianjin University of Technology, Tianjin 300384, China
3 shared publications Guilian Wang
Tianjin Key Laboratory for Advanced Mechatronic System Design and Intelligent Control, School of Mechanical Engineering, Tianjin University of Technology, Tianjin, China
2 shared publicationsHaibo Zhou
Tianjin Key Laboratory for Advanced Mechatronic System Design and Intelligent Control, School of Mechanical Engineering, Tianjin University of Technology, Tianjin, China
2 shared publications- Haibo Zhou
Tianjin Key Laboratory for Advanced Mechatronic System Design and Intelligent Control, School of Mechanical Engineering, Tianjin University of Technology, Tianjin, China
2 shared publications
Latest Publications
- 5 June 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Medical Robotics and Bionics
- Vol. PP (3) , 1
Correlation Mechanism of Cranial Bone Mechanical Characteristics and CT Imaging Features in Robot-Assisted Craniotomy
In robot-assisted neurosurgical craniotomy, cranial drilling serves as a core step in establishing precise surgical access. However, the limited surgical field restricts direct observation of the drill-bone interface interaction dynamics. Existing systems lack the capability to predict mechanical responses based on preoperative imaging, which constrains the implementation of intelligent adaptive robotic drilling. In this paper, we developed a craniotomy robotic platform integrating a six-axis force/torque sensor and an optical navigation system. Aiming at the cranial drilling scenario in craniotomy surgery, through 1006 drilling experiments at multiple sites on 74 ex vivo animal skulls, we systematically investigated the quantitative mapping relationship between CT HU values and drilling forces/torques. A predictive model based on quadratic polynomial regression was established, achieving R2 = 0.852 for the drilling force model and R2 = 0.816 for the torque model, significantly outperforming both linear and power-law models. Cross-species validation experiments demonstrated robust generalization performance of the model across bovine and ovine skulls, with prediction errors below 6% in high-density regions. This study realized mechanical sample generalization verification at multiple anatomical sites of the skull for craniotomy drilling, providing experimental evidence and methodological support for preoperative mechanical prediction and adaptive force control of craniotomy surgical robots, and can improve the safety and intelligence level of surgery.
- 5 June 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Medical Robotics and Bionics
- Vol. PP (3) , 1
Correlation Mechanism of Cranial Bone Mechanical Characteristics and CT Imaging Features in Robot-Assisted Craniotomy
In robot-assisted neurosurgical craniotomy, cranial drilling serves as a core step in establishing precise surgical access. However, the limited surgical field restricts direct observation of the drill-bone interface interaction dynamics. Existing systems lack the capability to predict mechanical responses based on preoperative imaging, which constrains the implementation of intelligent adaptive robotic drilling. In this paper, we developed a craniotomy robotic platform integrating a six-axis force/torque sensor and an optical navigation system. Aiming at the cranial drilling scenario in craniotomy surgery, through 1006 drilling experiments at multiple sites on 74 ex vivo animal skulls, we systematically investigated the quantitative mapping relationship between CT HU values and drilling forces/torques. A predictive model based on quadratic polynomial regression was established, achieving R2 = 0.852 for the drilling force model and R2 = 0.816 for the torque model, significantly outperforming both linear and power-law models. Cross-species validation experiments demonstrated robust generalization performance of the model across bovine and ovine skulls, with prediction errors below 6% in high-density regions. This study realized mechanical sample generalization verification at multiple anatomical sites of the skull for craniotomy drilling, providing experimental evidence and methodological support for preoperative mechanical prediction and adaptive force control of craniotomy surgical robots, and can improve the safety and intelligence level of surgery.
- 8 December 2025
- Published by Elsevier in Pattern Recognition
- Vol. 173
Attention-driven refinement network for continuity-preserving airway segmentation in class-imbalanced CT
- 8 December 2025
- Published by Elsevier in Pattern Recognition
- Vol. 173
Attention-driven refinement network for continuity-preserving airway segmentation in class-imbalanced CT
- 17 October 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Human-Machine Systems
- Vol. 55 (6) , 953-962
A Neurosurgical Craniotomy Training System Based on Haptic Virtual Reality Simulation
Traditional neurosurgical training modes face challenges including high costs, limited resources, lengthy learning curves, and difficulties in personalized training. In this article, we developed an immersive neurosurgical craniotomy virtual training system (NeuroSimulator) that integrates haptic feedback, enabling comprehensive surgical skill learning through an operator-control interface. Specifically, we constructed the comprehensive neurosurgical craniotomy surgical procedural (CNCSP) dataset to guide operators in repetitive learning and personalized training of relevant surgical skills. To address surgical site model rendering complexity, we proposed an algorithm that integrates vertex curvature and edge-length cost calculation factors (VC&ECL-QEM), resolving the incompatibility between surgical area image rendering quality and efficiency. For intracranial soft tissue haptic deformation, we developed a hybrid soft tissue haptic deformation (HBD) model that combines mass-spring and volumetric elements, addressing the collapse and distortion issues of traditional models and achieving more realistic soft tissue haptic deformation. Experimental results demonstrate that VC&ECL-QEM can simplify nonsurgical area feature preservation while maintaining surgical site detail features, reflecting the effectiveness of model simplification. The HBD model focuses on improving soft tissue deformation realism and shows high consistency with finite element model deformation effects. A total of 83 participants highly recognized NeuroSimulator’s system performance in terms of operational compliance, rendering real-time performance, and deformation realism, achieving effective improvements in skill proficiency metrics including operation time, ineffective operations, guidance requests, and operation scores. NeuroSimulator provides an innovative, efficient, and practical solution for neurosurgical training and is expected to play an increasingly important role in medical education and clinical skill enhancement.
- 17 October 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Human-Machine Systems
- Vol. 55 (6) , 953-962
A Neurosurgical Craniotomy Training System Based on Haptic Virtual Reality Simulation
Traditional neurosurgical training modes face challenges including high costs, limited resources, lengthy learning curves, and difficulties in personalized training. In this article, we developed an immersive neurosurgical craniotomy virtual training system (NeuroSimulator) that integrates haptic feedback, enabling comprehensive surgical skill learning through an operator-control interface. Specifically, we constructed the comprehensive neurosurgical craniotomy surgical procedural (CNCSP) dataset to guide operators in repetitive learning and personalized training of relevant surgical skills. To address surgical site model rendering complexity, we proposed an algorithm that integrates vertex curvature and edge-length cost calculation factors (VC&ECL-QEM), resolving the incompatibility between surgical area image rendering quality and efficiency. For intracranial soft tissue haptic deformation, we developed a hybrid soft tissue haptic deformation (HBD) model that combines mass-spring and volumetric elements, addressing the collapse and distortion issues of traditional models and achieving more realistic soft tissue haptic deformation. Experimental results demonstrate that VC&ECL-QEM can simplify nonsurgical area feature preservation while maintaining surgical site detail features, reflecting the effectiveness of model simplification. The HBD model focuses on improving soft tissue deformation realism and shows high consistency with finite element model deformation effects. A total of 83 participants highly recognized NeuroSimulator’s system performance in terms of operational compliance, rendering real-time performance, and deformation realism, achieving effective improvements in skill proficiency metrics including operation time, ineffective operations, guidance requests, and operation scores. NeuroSimulator provides an innovative, efficient, and practical solution for neurosurgical training and is expected to play an increasingly important role in medical education and clinical skill enhancement.
- 9 May 2025
- Published by Springer Nature in Signal, Image and Video Processing
- Vol. 19 (7) , 1-11
Target-background feature blocks and aberrance repressed correlation filters for real-time UAV tracking
- 9 May 2025
- Published by Springer Nature in Signal, Image and Video Processing
- Vol. 19 (7) , 1-11
Target-background feature blocks and aberrance repressed correlation filters for real-time UAV tracking
- 3 April 2025
- Published by Elsevier in Physica Medica
- Vol. 133, 104968
Clinically applicable semi-supervised learning framework for multiple organs at risk and tumor delineation in lung cancer brachytherapy
- 3 April 2025
- Published by Elsevier in Physica Medica
- Vol. 133, 104968
Clinically applicable semi-supervised learning framework for multiple organs at risk and tumor delineation in lung cancer brachytherapy
- 27 February 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Instrumentation and Measurement
- Vol. 74 (00189456) , 1-16
Multidimensional Collaborative Attention-Driven Neural Network for Real-Time Neurosurgical Instrument Segmentation
The automatic identification and precise marking of neurosurgical instruments can assist physicians in optimizing their visual search and judgment processes, alleviating visual fatigue, and enhancing decision-making efficiency, enabling doctors to focus more on the precise operation process of intracranial surgery. However, limited field of view in the surgical area and complex collaboration among multiple instruments can lead to instrument jitter, local occlusion, insufficient lighting, smoke interference, and other factors, making it difficult to maintain the robustness and stability of instrument segmentation. This article proposes a lightweight neurosurgical multi-instrument real-time segmentation network driven by multidimensional collaborative attention, utilizing YOLOv8n as the backbone. We developed the concatenate-triplet attention (Cat-TA) module, which enhances global information exchange and integrates multidimensional attention to strengthen the representation of effective features of different instruments. The optimized omni-dimensional dynamic convolution C2f (ODC2f) module combines lightness with the capability of multiscale precise perception, better adapting to the online recognition and segmentation tasks for various instruments. Additionally, we employ the Wise-IoUv3 (WIoUv3) for the boundary box adjustment error function, with a variable incremental nonlinear mechanism to optimize the gradient distribution strategy, further improving the overall segmentation performance. The robustness and efficiency of the proposed method were proven by a dataset of surgical instruments for intracranial gliomas collaboratively created with professional physicians, which achieved an Mean Pixel Accuracy (mPA) of 95.2% and an Mean Intersection over Union (MIoU) of 89.8%. Extensive experiments demonstrated that our proposed method exhibits excellent performance in understanding surgical scenarios, increasing the safety and success rate of operations.
- 27 February 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Instrumentation and Measurement
- Vol. 74 (00189456) , 1-16
Multidimensional Collaborative Attention-Driven Neural Network for Real-Time Neurosurgical Instrument Segmentation
The automatic identification and precise marking of neurosurgical instruments can assist physicians in optimizing their visual search and judgment processes, alleviating visual fatigue, and enhancing decision-making efficiency, enabling doctors to focus more on the precise operation process of intracranial surgery. However, limited field of view in the surgical area and complex collaboration among multiple instruments can lead to instrument jitter, local occlusion, insufficient lighting, smoke interference, and other factors, making it difficult to maintain the robustness and stability of instrument segmentation. This article proposes a lightweight neurosurgical multi-instrument real-time segmentation network driven by multidimensional collaborative attention, utilizing YOLOv8n as the backbone. We developed the concatenate-triplet attention (Cat-TA) module, which enhances global information exchange and integrates multidimensional attention to strengthen the representation of effective features of different instruments. The optimized omni-dimensional dynamic convolution C2f (ODC2f) module combines lightness with the capability of multiscale precise perception, better adapting to the online recognition and segmentation tasks for various instruments. Additionally, we employ the Wise-IoUv3 (WIoUv3) for the boundary box adjustment error function, with a variable incremental nonlinear mechanism to optimize the gradient distribution strategy, further improving the overall segmentation performance. The robustness and efficiency of the proposed method were proven by a dataset of surgical instruments for intracranial gliomas collaboratively created with professional physicians, which achieved an Mean Pixel Accuracy (mPA) of 95.2% and an Mean Intersection over Union (MIoU) of 89.8%. Extensive experiments demonstrated that our proposed method exhibits excellent performance in understanding surgical scenarios, increasing the safety and success rate of operations.
- Open Access
- 22 November 2024
- Published by Association for Computing Machinery (ACM)
- p. 250-255
Project-guided Teaching Reform of Image Processing and Machine Vision: A Case Study of Intelligent Diagnosis of Pulmonary Nodules
- 22 November 2024
- Published by Association for Computing Machinery (ACM)
- p. 250-255
Open AccessProject-guided Teaching Reform of Image Processing and Machine Vision: A Case Study of Intelligent Diagnosis of Pulmonary Nodules
- Open Access
- 8 November 2024
- Published by Association for Computing Machinery (ACM)
- p. 509-514
Innovative Application of Project Integration Technology in the Teaching Mode of Engineering Drawing Course
Taking underwater robots project as an example, this paper discusses the innovation of the teaching mode of engineering drawing courses under the guidance of a project. The traditional engineering drawing course often lacks the link of practice and project application, which leads to the lack of students’ understanding and application ability of the course content. Therefore, this study proposes a teaching mode based on project guidance. By using underwater robots as teaching cases, theoretical knowledge and practical application are combined to promote the cultivation of students’ comprehensive abilities. This paper first reviews the research progress in related fields, then introduces the design and implementation methods of the teaching mode in detail, and analyzes the actual effect of the model through the underwater robot case. The results show that the teaching mode of engineering drawing courses under the guidance of the project has a positive impact on students ' learning interest, practical ability and teamwork ability. The results of this study will provide practical and feasible solutions for the innovation of the teaching mode of engineering drawing courses, and also provide a reference for the reform of undergraduate teaching mode in the construction of “New Engineering Education”.
- 8 November 2024
- Published by Association for Computing Machinery (ACM)
- p. 509-514
Open AccessInnovative Application of Project Integration Technology in the Teaching Mode of Engineering Drawing Course
Taking underwater robots project as an example, this paper discusses the innovation of the teaching mode of engineering drawing courses under the guidance of a project. The traditional engineering drawing course often lacks the link of practice and project application, which leads to the lack of students’ understanding and application ability of the course content. Therefore, this study proposes a teaching mode based on project guidance. By using underwater robots as teaching cases, theoretical knowledge and practical application are combined to promote the cultivation of students’ comprehensive abilities. This paper first reviews the research progress in related fields, then introduces the design and implementation methods of the teaching mode in detail, and analyzes the actual effect of the model through the underwater robot case. The results show that the teaching mode of engineering drawing courses under the guidance of the project has a positive impact on students ' learning interest, practical ability and teamwork ability. The results of this study will provide practical and feasible solutions for the innovation of the teaching mode of engineering drawing courses, and also provide a reference for the reform of undergraduate teaching mode in the construction of “New Engineering Education”.