Label-free intraoperative histological imaging

点击翻译按钮获取中文摘要

The aim of this project is to apply machine learning to histological imaging to determine the difference between cancerous and healthy tissue. At the moment, when a surgeon performs surgery to remove tumour tissue, histological samples are taken and reviewed separately in a pathology lab by experts with the results returned to the surgeon up to two weeks later. This can result in the patient needing a second surgery to remove any remaining cancerous tissue, something that is dangerous and costly. This project aims to apply machine learning so that when a histological sample is taken, it can be placed under a microscope and analysed by an artificial neural network during the initial surgery and results provided whilst the patient is still under general anaesthesia. This technique will then be applied to the Mesolens to provide a larger field of imaging and greater analysis of the sample. Training will take place at Strathclyde University as well as the National Physical Laboratory and University College London to cover the hardware and software aspects of this project.