Measurement Synchronization for Fluorescence Lifetime Imaging using Image Recognition for In-Vivo Imaging: ■
Subject ■
With the unique ETRO CAPS nanosecond gated camera [1], it is possible to image the nanosecond time behavior of fluorescence: the fluorescence lifetime (figure 1). This enables tumors which have been marked with fluorescent contrast agents to be imaged with better contrast. ETRO has built a complete imaging system, consisting of such a CAPS camera and a laser to excite the fluorescence and this system is currently in use for cancer research on mice and dogs.
During these in vivo experiments, it has been observed that the acquired measurements are often affected by motion artefacts originating from the subject. These artefacts are mainly caused by physiological movements, such as breathing, and can significantly distort the measured temporal fluorescence signals.
The objective of this thesis is to investigate image recognition-based methods to compensate for these motion-induced artefacts and to enable accurate reconstruction of the fluorescence time behavior.
Motion information can be obtained using the integrated color camera (part of the gated camera system), by detecting and tracking the breathing cycle. This information can then be used to guide the reconstruction process. Several reconstruction approaches can be considered. These include methods based on analytical models, as well as data-driven approaches such as deep learning.
A key challenge is the co-registration of the CAPS gated sensor and the color sensor, using either the integrated distance sensor or feature-based alignment. Additionally, for ex vivo specimens, the fluorescence and color images can be co-registered on a 3D map obtained from tabletop 3D scanner (e.g. einstar 3D scanner).
During these in vivo experiments, it has been observed that the acquired measurements are often affected by motion artefacts originating from the subject. These artefacts are mainly caused by physiological movements, such as breathing, and can significantly distort the measured temporal fluorescence signals.
The objective of this thesis is to investigate image recognition-based methods to compensate for these motion-induced artefacts and to enable accurate reconstruction of the fluorescence time behavior.
Motion information can be obtained using the integrated color camera (part of the gated camera system), by detecting and tracking the breathing cycle. This information can then be used to guide the reconstruction process. Several reconstruction approaches can be considered. These include methods based on analytical models, as well as data-driven approaches such as deep learning.
A key challenge is the co-registration of the CAPS gated sensor and the color sensor, using either the integrated distance sensor or feature-based alignment. Additionally, for ex vivo specimens, the fluorescence and color images can be co-registered on a 3D map obtained from tabletop 3D scanner (e.g. einstar 3D scanner).
Kind of work ■
The objective of this thesis is to investigate image recognition-based methods to compensate for these motion-induced artefacts and to enable accurate reconstruction of the fluorescence time behavior.
Specific objectives are to:
(1)
Identify existing approaches for motion correction and image registration.
(2)
Tackling the differences in sensor resolution and position and ensuring a robust mapping.
(3)
Implement the motion correction to acquire fluorescence lifetime images.
(4)
Create an overlay of the fluorescence lifetime image on 3D images.
Specific objectives are to:
(1)
Identify existing approaches for motion correction and image registration.
(2)
Tackling the differences in sensor resolution and position and ensuring a robust mapping.
(3)
Implement the motion correction to acquire fluorescence lifetime images.
(4)
Create an overlay of the fluorescence lifetime image on 3D images.
Framework of the Thesis ■
The project will consist of:
•
A literature review to identify existing approaches for motion correction and image registration.
•
An analysis of the available datasets that can be used for testing and validating the identified metrics.
•
Implementing the most relevant methods on the ETRO imaging system, taking into account the differences in sensor resolution.
•
Validating the approach in relevant experiments and comparing the results to conventional measurements.
•
Optionally, adding a 3D reconstruction of the tissue with fluorescence lifetime overlay.
•
A literature review to identify existing approaches for motion correction and image registration.
•
An analysis of the available datasets that can be used for testing and validating the identified metrics.
•
Implementing the most relevant methods on the ETRO imaging system, taking into account the differences in sensor resolution.
•
Validating the approach in relevant experiments and comparing the results to conventional measurements.
•
Optionally, adding a 3D reconstruction of the tissue with fluorescence lifetime overlay.