Real-time tracking and classification of tumour and non-tumour tissue in upper gastrointestinal cancers using diffuse reflectance spectroscopy for resection margin assessment
File(s)
Published version
Author(s)
Type
Journal Article
Abstract
Importance:
Cancers of the upper gastrointestinal tract remain a major contributor to the global cancer burden. The accurate mapping of tumour margins is of particular importance for curative cancer resection and improvement in overall survival. Current mapping techniques preclude a full resection margin assessment in real-time.
Objective:
We aimed to use diffuse reflectance spectroscopy on gastric and oesophageal cancer specimens to differentiate tissue types and provide real-time feedback to the operator.
Design:
This was a prospective ex vivo validation study. Patients undergoing oesophageal or gastric cancer resection were prospectively recruited into the study between July 2020 and July 2021 at Hammersmith Hospital in London, United Kingdom.
Setting:
This was a single-centre study based at a tertiary hospital.
Participants:
Tissue specimens were included for patients undergoing elective surgery for either oesophageal carcinoma (adenocarcinoma or squamous cell carcinoma) or gastric adenocarcinoma.
Exposure:
A hand-held diffuse reflectance spectroscopy probe and tracking system was used on freshly resected ex vivo tissue to obtain spectral data. Binary classification, following histopathological validation, was performed using four supervised machine learning classifiers.
Main Outcomes and Measures:
Data were divided into training and testing sets using a stratified 5-fold cross-validation method. Machine learning classifiers were evaluated in terms of sensitivity, specificity, overall accuracy, and the area under the curve.
Results:
A total of 14,097 mean spectra for normal and cancerous tissue were collected from 37 patients. The machine learning classifier achieved an overall normal versus cancer diagnostic accuracy of 93.86±0.66 for stomach tissue and 96.22±0.50 for oesophageal tissue, and sensitivity and specificity of 91.31% and 95.13% for stomach and 94.60% and 97.28% for oesophagus, respectively. Real-time tissue tracking and classification was achieved and presented live on-screen.
Conclusion:
This study provides ex vivo validation of the diffuse reflectance spectroscopy technology for real-time differentiation of gastric and oesophageal cancer from healthy tissue using machine learning with high accuracy. As such, it is a step towards the development of a real-time in vivo tumour mapping tool for oesophageal and gastric cancers, that can aid decision-making of resection margins intra-operatively.
Cancers of the upper gastrointestinal tract remain a major contributor to the global cancer burden. The accurate mapping of tumour margins is of particular importance for curative cancer resection and improvement in overall survival. Current mapping techniques preclude a full resection margin assessment in real-time.
Objective:
We aimed to use diffuse reflectance spectroscopy on gastric and oesophageal cancer specimens to differentiate tissue types and provide real-time feedback to the operator.
Design:
This was a prospective ex vivo validation study. Patients undergoing oesophageal or gastric cancer resection were prospectively recruited into the study between July 2020 and July 2021 at Hammersmith Hospital in London, United Kingdom.
Setting:
This was a single-centre study based at a tertiary hospital.
Participants:
Tissue specimens were included for patients undergoing elective surgery for either oesophageal carcinoma (adenocarcinoma or squamous cell carcinoma) or gastric adenocarcinoma.
Exposure:
A hand-held diffuse reflectance spectroscopy probe and tracking system was used on freshly resected ex vivo tissue to obtain spectral data. Binary classification, following histopathological validation, was performed using four supervised machine learning classifiers.
Main Outcomes and Measures:
Data were divided into training and testing sets using a stratified 5-fold cross-validation method. Machine learning classifiers were evaluated in terms of sensitivity, specificity, overall accuracy, and the area under the curve.
Results:
A total of 14,097 mean spectra for normal and cancerous tissue were collected from 37 patients. The machine learning classifier achieved an overall normal versus cancer diagnostic accuracy of 93.86±0.66 for stomach tissue and 96.22±0.50 for oesophageal tissue, and sensitivity and specificity of 91.31% and 95.13% for stomach and 94.60% and 97.28% for oesophagus, respectively. Real-time tissue tracking and classification was achieved and presented live on-screen.
Conclusion:
This study provides ex vivo validation of the diffuse reflectance spectroscopy technology for real-time differentiation of gastric and oesophageal cancer from healthy tissue using machine learning with high accuracy. As such, it is a step towards the development of a real-time in vivo tumour mapping tool for oesophageal and gastric cancers, that can aid decision-making of resection margins intra-operatively.
Date Issued
2022-09-07
Date Acceptance
2022-06-23
Citation
JAMA Surgery, 2022, 157 (11)
ISSN
2168-6254
Publisher
American Medical Association
Journal / Book Title
JAMA Surgery
Volume
157
Issue
11
Copyright Statement
© 2022 Nazarian S et al. This is an open access article distributed under the terms of the CC-BY License (https://creativecommons.org/licenses/by/4.0/)
License URL
Identifier
10.1001/jamasurg.2022.3899
Publication Status
Published
Article Number
e223899
Date Publish Online
2022-09-07