Modelling, Simulation and Control of Needle Insertion in Soft Tissues with Application to Transperineal Prostate Biopsy.

  • Athanasios Martsopoulos

Student thesis: Doctoral ThesisDoctor of Philosophy (PhD)

Abstract

Transperineal prostate biopsy constitutes one of the key tools for prostate cancer diagnosis
and classification. The success of the biopsy is highly dependent on the accuracy of the needle
placement, with imprecise targeting often leading to severe complications, such as false negatives
or ablation of healthy tissue. However, accurate manual percutaneous needle placement is a
highly challenging task due to the unknown and complex properties of needle-tissue interaction.
Robot-assisted prostate biopsy and virtual surgical training platforms have the potential to
significantly improve the accuracy of these operations.
The aim of this Thesis is to investigate, design and implement mathematical models,
experimental studies, and control algorithms that will enable improvements in the modelling,
simulation and control of needle insertions for transperineal prostate biopsy procedures. A
systematic review of existing simulation and robotic solutions is conducted to understand the
current needs in both surgical training and robot-assisted needle insertion. The background
research shows that there is a need for physically accurate and computationally efficient
representations of the underlying mechanics of flexible needles and soft tissues, as well as for the
dynamics of their interaction. In addition, robot-assisted needle insertion systems require control
architectures that prioritise safety and are able to adapt to the unknown and dynamically
changing properties of human tissue and needle-tissue interaction.
In this regard, a number of novel methodologies, in both needle and tissue modelling, are
presented, aiming to provide both physically accurate and computationally efficient simulation
solutions, without the need for simplifying modelling assumptions. The needle and tissue models
are then combined to provide a computationally efficient constraint-based model that aims
to accurately capture the underlying mechanics of needle-tissue interaction. The presented
mathematical models are also combined for the implementation of a novel needle insertion
simulator. In addition, the Thesis presents the design and implementation of an adaptive and
safety-focused control system for robot-assisted and highly accurate needle insertions in soft
tissue. A set of user studies in manual needle insertions is also conducted to identify manual
needle guidance strategies, inform the design of early-stage training methodologies in needle
insertion and shape future developments in robot-assisted insertion systems. The presented
results are systematically validated through specially designed experimental studies.
Date of Award18 Jun 2024
Original languageEnglish
Awarding Institution
  • University of Bristol
SupervisorAntonia Tzemanaki (Supervisor) & Tom L Hill (Supervisor)

Cite this

'