Coventry University Featured PhD Programmes
University of Huddersfield Featured PhD Programmes
University of Edinburgh Featured PhD Programmes
University of Glasgow Featured PhD Programmes
Karlsruhe Institute of Technology Featured PhD Programmes

PhD Project - Hyperspectral sensing and imaging device for gastrointestinal cancer surgery


About This PhD Project

Project Description

Multispectral optical probes promise sensitivity and specificity above 90% for discriminating between normal and abnormal tissue. One of these techniques, diffuse reflectance spectroscopy (DRS), utilises a completely safe broadband light source, optical fibres for light delivery and collection, a spectrograph to disperse the emitted light into the different wavelengths, and a detector (commonly a CCD camera). When compared to sophisticated microendoscopic probes, DRS has lower costs, and is simpler because it does not require lasers or magnification optics. Furthermore, combining DRS with fluorescence spectroscopy can offer increased accuracy through a dual modality approach that nevertheless maintains a simple probe design for clinical deployment.

We propose the use of a miniaturised endoscopic 2-D array of single-point DRS probe sensors for hyperspectral dense circumferential resection margin (CRM) assessment, based on technology developed in the Elson group. Such a device could be used directly in the surgical field, allowing the discrimination of tumour from scar tissue in real-time, minimising the time and cost of the CRM assessment, and optimising the clinical outcomes for patients with GI cancer. Furthermore, we propose to adapt this approach to transform it from a single point spectroscopic method to an imaging device that is suitable for in vivo data acquisition.

We are seeking a PhD student with strong engineering or physical sciences background with an interest in applying their skills to the design, construction, testing and information analysis of a novel surgical device.

Aim of this study

Is it possible to analyse intraoperative circumferential resection margins by using hyperspectral probes for automated tissue classification?

Rationale

The 5-year survival rate for patients with locoregional advanced carcinoma is below 40%. After chemoradiotherapy, the tumour volume is reduced, but an outer core of scar tissue is created. Even using preoperative imaging (CT and MRI), it is hard to differentiate tumour from scar tissue. Positive (CRM) is a predictor of local recurrence of GI cancers but identifying CRMs in the theatre is a challenge because tumour and non-tumour scar tissue are macroscopically very similar. The current technique involves sending small samples of the area of interest off for histopathological analysis using frozen sections. This is limited by taking at least 30 minutes to process and there are only a finite number of areas that can be sampled. It also leads to a significant increase in operative time affecting both the patient recovery and theatre efficiency.

We have tested the principles of DRS tissue discrimination, where fresh ex vivo cancer samples were scanned with a hand-held DRS probe. A total of 30,600 spectra were collected from 179 ex vivo samples. DRS spectra were then matched with the histopathology report. In this pilot study, normal and cancerous tissue had statistically significant differences in DRS intensities across the spectrum (p<0.01, e.g. figure 1), and normal tissue post chemoradiation had significantly different spectra compared to treatment naive tissue (p<0.01).

We now need to transition this device to being an intraoperative sterilisable tool and have begun this process by adapting a standard surgical laparoscope. Once we can collect intraoperative data we can begin the process creating a database of spectra, with matched histopathological diagnosis, and then commence machine learning to allow real time tissue classification. Once we have achieved a robust classification performance we can then begin the validation of the technology. The super-spectral-imaging approach has great promise and the feasibility has been demonstrated and published, but there are a number of technical/computing challenges to solve for its clinical application and it should be validated through the interaction with the surgical teams led by Mr Peters.

The supervisory team cover surgery and physical sciences/engineering disciplines, with the overall concept designed by the team as a whole. The student would initially have a desk at Imperial, South Kensington Campus and a hot desk at St Mary’s Hospital. Development activities would be complemented by meetings every two weeks with the clinical team.

Email Now

Insert previous message below for editing? 
You haven’t included a message. Providing a specific message means universities will take your enquiry more seriously and helps them provide the information you need.
Why not add a message here
* required field
Send a copy to me for my own records.

Your enquiry has been emailed successfully





FindAPhD. Copyright 2005-2019
All rights reserved.