Temporally-multiplexed machine-vision based fluorescence imaging for image-guided resection of primary brain tumours

2020 Research Internship Project


Faculty Name

Victor Yang

Project Title

Temporally-multiplexed machine-vision based fluorescence imaging for image-guided resection of primary brain tumours

Project Description

Fluorescence guided surgical resection has held great promise as a tool for resection guidance in various surgeries, including for treatment of primary brain tumours. These systems often employ various fluorophores that see preferential uptake for pathological tissue, allowing them to be highlighted and visualized when illuminated by the excitation source. This has been seen in particular with 5-ALA/PpIX and ICG based resections for glioma and arteriovenous malformations, respectively. While this does improve differentiation of healthy and diseased tissue, the systems are often rudimentary, consisting of a basic excitation source and visualization system that requires all lighting in the operating room to be off, disrupting OR staff workflow and adding operation time. Further, interpretation of the fluorescence signal is subjective and up to the interpretation of the surgeon in order to determine the areas that are and are not fluorescing.

For this reason, we propose the creation of a temporally-multiplexed machine-vision based fluorescence imaging system. In this system, all room lighting will be synchronized with illumination and excitation sources, allowing for the lights to go through synchronized "light" and "dark" periods at frequencies beyond the perceptibility of humans, but which can be gated to allow for camera imaging during either light or dark phases. This would allow for fluorescence imaging to occur at the dark phases without disruption to the OR workflow, and may allow for higher sensitivity imaging modalities to be used. Moreover, images will be obtained using the machine-vision cameras equipped with band-pass filters to detect only emission wavelengths, and will be processed using various image-processing and computer vision techniques to segment and overlay fluorescence signals onto a white-light camera image, enabling quantitative differentiation of fluorescence signals. This will be evaluated in a series of phantom and in vivo models.

Student Responsibility

The student will work closely with two graduate researchers on numerous aspects of the project. The student will, under the guidance of the graduate researchers, be responsible for the construction and verification of the phantoms for testing, as well as assist in the phantom validation process. The student will also be responsible for assisting in the development of the live-feed processing pipeline, and will assist in carrying out various user and validation tests needed for the project. In this, the student will have consistent and regular interaction with engineers, researchers, clinicians, and clinician-scientists, who will be guiding the design and testing of the system throughout this process.

Specific Requirements

- Student should have a good working familiarity with MATLAB - Experience in Python and C++, as well as image-processing experience is preferred, but not necessary - Experience in CAD, 3D printing, and modelling with materials and polymers is an asset - Student should be able to work independently and in a team - Experience in human factors considerations also an asset

Reseach Internship Application

Victor Yang : Temporally-multiplexed machine-vision based fluorescence imaging for image-guided resection of primary brain tumours | Thursday March 12th 2020 10:29 AM