Related Experiment Video
Updated: Nov 15, 2025
07:22
Minimally Invasive Murine Laryngoscopy for Close-Up Imaging of Laryngeal Motion During Breathing and Swallowing
Published on: December 1, 2023
792
Target localization during respiration motion based on LSTM: A pilot study on robotic puncture system
Yuxiang Ma1,2, Zhikai Yang1,2, Wei Wu1,2
1School of Biomedical Engineering, Shanghai Jiao Tong University, Shanghai, China.
Summary
This study introduces a new method using long short-term memory (LSTM) to accurately track tumors during respiratory motion. The approach significantly improves localization accuracy for needle biopsies and robotic-assisted procedures.
Area of Science:
- Medical Imaging
- Robotics
- Artificial Intelligence
Background:
- Respiratory motion complicates accurate localization in needle biopsies.
- Existing methods struggle with accuracy due to complex internal tissue motion.
Purpose of the Study:
- To improve target localization accuracy during respiratory motion.
- To reduce the complexity of predictive models for respiratory compensation.
Main Methods:
- A novel framework using long short-term memory (LSTM) for target localization.
- Utilizing principal components of external surrogate signals to predict internal tumor trajectory.
- Integration with an electromagnetic tracking system and robotic arm for real-time tumor tracking.
Main Results:
- Achieved an average mean absolute error of 0.44 mm and root-mean-square error of 0.58 mm on public datasets.
- Demonstrated an average root mean square error of 0.65 mm in motion phantom studies.
- Validated the method's effectiveness in a prototype robotic puncture system.
Conclusions:
- The proposed LSTM-based method enhances target localization accuracy during respiratory movement.
- The approach shows significant potential for clinical applications in image-guided interventions.
