Computer-assisted intraoperative ultrasound in brain tumour surgery
File(s)
Author(s)
Weld, Alistair
Type
Thesis
Abstract
Intraoperative imaging has been shown to be a vital tool for surgical decision making, particularly in the case of brain tumour surgery. Real-time acquisition of subsurface information is essential for achieving maximal safe resection, removing as much pathological tissue as possible while minimising damage to normal and critical tissue. Ultrasound has shown considerable promise in becoming the principal tool for assisting neurosurgeons. However, despite this potential, due to the lack of standardisation of training and use, the variability between centres is high. This factor has, as a consequence, contributed to the lack of research on this imaging tool, especially compared to other similar modalities. Brain tumours are an especially complex type of tumour and the surgical process is likely to incur many confounders, making real-time subsurface imaging essential. In particular, in the case of diffuse tumours, where due to the inflitrative properties, the margins suffer from ill-definition, which can introduce significant uncertainty and subjectivity during surgical decision making.
This thesis presents research on computer-assisted methods for intraoperative ultrasound for brain tumour resection. The work was carried out as part of the Imperial–Technische Universität München (TUM) Joint Academy of Doctoral Studies, in collaboration with the Department of Neurosurgery and Imaging at Charing Cross Hospital. The work at Imperial, conducted at the Hamlyn Centre, primarily addressed computer vision topics relevant to the clinical context, while TUM and the Deutsches Zentrum für Luft- und Raumfahrt (German Aerospace Centre) focused on robotics and control systems.
The thesis contributions can be grouped into three categories: Pathological Tissue Delineation, Probe Manipulation Analysis, Resource Leveraging. Together, the work in these chapters aims to expand the understanding of ultrasound brain tumour imaging and to present solutions to enable real-time computer-assisted support for surgeons and facilitate automation through robotic systems.
This thesis presents research on computer-assisted methods for intraoperative ultrasound for brain tumour resection. The work was carried out as part of the Imperial–Technische Universität München (TUM) Joint Academy of Doctoral Studies, in collaboration with the Department of Neurosurgery and Imaging at Charing Cross Hospital. The work at Imperial, conducted at the Hamlyn Centre, primarily addressed computer vision topics relevant to the clinical context, while TUM and the Deutsches Zentrum für Luft- und Raumfahrt (German Aerospace Centre) focused on robotics and control systems.
The thesis contributions can be grouped into three categories: Pathological Tissue Delineation, Probe Manipulation Analysis, Resource Leveraging. Together, the work in these chapters aims to expand the understanding of ultrasound brain tumour imaging and to present solutions to enable real-time computer-assisted support for surgeons and facilitate automation through robotic systems.
Version
Open Access
Date Issued
2025-02-28
Date Awarded
01/12/2025
Advisor
Giannarou, Stamatia
Sponsor
UK Research and Innovation
Grant Number
EP/S023283/1
Publisher Department
Department of Computing
Publisher Institution
Imperial College London
Qualification Level
Doctoral
Qualification Name
Doctor of Philosophy (PhD)