# A New Tensioning Method using Deep Reinforcement Learning for Surgical Pattern Cutting

**Authors:** Thanh Thi Nguyen, Ngoc Duy Nguyen, Fernando Bello, Saeid Nahavandi  
**arXiv:** [1901.03327](https://arxiv.org/abs/1901.03327)  
**Date:** 2019-01-10  
**Summary page:** https://www.summarizepaper.com/en/paper/1901.03327v1/

## Key points

- New tensioning method for surgical pattern cutting using deep reinforcement learning
- Utilizes a multiple pinch point approach to model an autonomous tensioning planner
- Outperforms existing methods in terms of performance and robustness
- Authors: Thanh Thi Nguyen, Ngoc Duy Nguyen, Fernando Bello, and Saeid Nahavandi
- Presented at the 2019 IEEE International Conference on Industrial Technology (ICIT) in Melbourne, Australia
- Surgeons rely on surgical scissors and tissue grippers for cutting deformable surgical tissue
- Cutting accuracy depends on skills in manipulating tools
- Comprehensive tensioning policy is important for improving cutting accuracy and reliability
- Automating tensioning policy benefits human surgeons and surgical robots
- Deep reinforcement learning algorithms can improve surgical pattern cutting techniques

## Summary

This paper presents a new tensioning method for surgical pattern cutting using deep reinforcement learning. The proposed method utilizes a multiple pinch point approach to model an autonomous tensioning planner based on a deep reinforcement learning algorithm. By training the algorithm on simulated scenarios, experiments show that this approach outperforms existing methods in terms of both performance and robustness. The authors of this paper are Thanh Thi Nguyen, Ngoc Duy Nguyen, Fernando Bello, and Saeid Nahavandi. The research was presented at the 2019 IEEE International Conference on Industrial Technology (ICIT) in Melbourne, Australia. Surgeons often rely on surgical scissors and tissue grippers to cut through deformable surgical tissue, and the accuracy of the cutting process depends on their skills in manipulating these tools. To improve cutting accuracy and reliability, it is important to have a comprehensive tensioning policy that determines the appropriate tensioning direction at each step of the cutting process. Automating this tensioning policy can benefit both human surgeons and surgical robots. The paper provides valuable insights into improving surgical pattern cutting techniques by leveraging deep reinforcement learning algorithms. It demonstrates how such algorithms can be used to develop an autonomous tensioning planner for improved performance and robustness in surgical pattern cutting processes.