Hyperspectral circumferential resection margin assessment for gastrointestinal cancer surgery
File(s)
Author(s)
Gkouzionis, Ioannis A
Avila-Rencoret, Fernando
Peters, Christopher
Elson, Daniel
Type
Thesis
Abstract
Gastrointestinal (GI) cancers represent a significant contributor to the global cancer burden. Surgery typically constitutes the primary therapeutic approach for GI malignancies, with the primary objective of maximising tumour tissue excision while maintaining an adequate margin of healthy tissue (clear margin) surrounding the tumour. The presence of a clear margin, devoid of cancer cells, is essential in minimising the risk of cancer recurrence. Accurate mapping of tumour margins is crucial for achieving a negative resection margin. Local recurrence following surgical resection significantly impacts the overall survival and quality of life for patients with GI cancer.
In this thesis, we have investigated diffuse reflectance spectroscopy (DRS) technology, combined with artificial intelligence, to develop a pioneering optical surgical tool aimed at improving the identification of resection margins in GI cancer surgery.
The primary goal is to enhance the precision and reliability of tumour margin demarcation, thereby increasing the likelihood of complete tumour removal with clear margins, which is critical for reducing the risk of cancer recurrence and improving patient survival and quality of life. This work presents a two-fold unique contribution towards the optimal resection margins assessment in GI surgery. Firstly, by introducing a novel exploratory data analysis and classification routine employing deep neural networks to distinguish healthy and malignant tissues based on spectral measurements, and secondly, by developing a deep learning model to facilitate the detection and tracking of the DRS fibre probe during surgery. This approach represents a significant advancement over previous studies by providing a comprehensive analysis using machine learning techniques for GI cancer discrimination and addressing the challenge of probe detection and tracking for the first time.
Through validation in clinical settings, this work aims to advance the standard of care in GI cancer surgery, demonstrating the potential of DRS technology and AI to improve surgical outcomes.
In this thesis, we have investigated diffuse reflectance spectroscopy (DRS) technology, combined with artificial intelligence, to develop a pioneering optical surgical tool aimed at improving the identification of resection margins in GI cancer surgery.
The primary goal is to enhance the precision and reliability of tumour margin demarcation, thereby increasing the likelihood of complete tumour removal with clear margins, which is critical for reducing the risk of cancer recurrence and improving patient survival and quality of life. This work presents a two-fold unique contribution towards the optimal resection margins assessment in GI surgery. Firstly, by introducing a novel exploratory data analysis and classification routine employing deep neural networks to distinguish healthy and malignant tissues based on spectral measurements, and secondly, by developing a deep learning model to facilitate the detection and tracking of the DRS fibre probe during surgery. This approach represents a significant advancement over previous studies by providing a comprehensive analysis using machine learning techniques for GI cancer discrimination and addressing the challenge of probe detection and tracking for the first time.
Through validation in clinical settings, this work aims to advance the standard of care in GI cancer surgery, demonstrating the potential of DRS technology and AI to improve surgical outcomes.
Version
Open Access
Date Issued
2023-09
Date Awarded
2024-03
Date Acceptance
2020-01-01
License URL
Advisor
Elson, Daniel
Peters, Christopher
Publisher Department
Surgery and Cancer
Publisher Institution
Imperial College London
Source
Biophotonics and Imaging Graduate Summer School 2020
Qualification Level
Doctoral
Qualification Name
Doctor of Philosophy (PhD)
Start Date
2020-08-25
Coverage Spatial
National University of Ireland Galway