Quantitative imaging of bioluminescent signals in freely behaving mice

The objective of this thesis is to design and validate algorithms to quantitatively detect bioluminescent signals simultaneously on several non-anesthetized mice. This thesis project is based on three axes which are :

1. To develop algorithms to correct dynamic 2D BLI signals from an imaging system of freely behaving mice and to improve signal localization. The quantification of these signals is altered by the different orientations of the mouse with respect to the detector or between them (if there are several animals) ;
2. To develop tools for automatic and unsupervised segmentation and tracking of several freely behaving mice ;
3. To test and validate the system and algorithm performances

It will be necessary to combine a profilometry technique allowing to extract quantitative information from the signal independently of the position of the animal with an imaging acquisition technique of freely behaving mice. For this purpose, we will develop algorithms based on data from dynamic profilometry, on mouse modeling based on physical models, and on the matching of the optical and infrared image by Machine Learning to follow the temporal variations of the BLI emission. This work will strengthen Biospace Lab’s leading position in the competitive market for real-time optical imaging of small animals.