Research Article
A Novel Visualization Method of Vessel Network for Tumour Targeting: A Vessel Matrix Approach
@INPROCEEDINGS{10.1007/978-3-031-43135-7_5, author={Mengsheng Zhai and Minghao Liu and Zhijing Wang and Yifan Chen and Yue Sun}, title={A Novel Visualization Method of Vessel Network for Tumour Targeting: A Vessel Matrix Approach}, proceedings={Bio-inspired Information and Communications Technologies. 14th EAI International Conference, BICT 2023, Okinawa, Japan, April 11-12, 2023, Proceedings}, proceedings_a={BICT}, year={2023}, month={9}, keywords={tumour targeting Vessel Network Visualization Reinforcement learning Vessel Matrix Model CONA}, doi={10.1007/978-3-031-43135-7_5} }- Mengsheng Zhai
Minghao Liu
Zhijing Wang
Yifan Chen
Yue Sun
Year: 2023
A Novel Visualization Method of Vessel Network for Tumour Targeting: A Vessel Matrix Approach
BICT
Springer
DOI: 10.1007/978-3-031-43135-7_5
Abstract
When the tumour grows, the microvessel density in the surrounding area increases and exhibits irregular curvature, which shows a difference from the regular vascular network. Therefore, a model is needed to describe the vascular network around the tumour. However, the existing models can not provide a good representation of the vascular network. This paper proposes a Vessel Matrix Model (VMM), a visualization vascular network model which has the potential to resemble the complicated vessels networks around the tumour microenvironment. VMM is conducive to the works such as drug delivery and tumour search and can perform a tumour-targeting search by combining with the computational nanobiosensing (CONA) framework. CONA uses nanorobots as computing agents to learn the surrounding environment to regulate the path-planning to the tumour location through algorithms such as reinforcement learning. A CONA method is performed in searching for a tumour to verify the feasibility of this vascular network. In order to seek optimal routing in the vascular network, VMM provides distance reward and weight reward for the agents, where the rewards are determined by the distance of starting point to the tumour lesion and the gradient of BGF, respectively. Therefore, VMM enables the tumour search with the CONA method. By introducing different weights between the destination and weights rewards, it is found that targeting efficiency can be affected by branch rate and size of the network.
