Autonomous High-Throughput accurate FLIM with subcellular resolution
Development and optimization of automated data mining from multi-well microscopy plates. The goal is to improve on current intensity-based methods for detection of the presence of biological phenomena in microscopic images by applying deep-learning for morphological analysis in detection and classification. Additional discriminating dimensions are created by integrating Fluorescence Lifetime Imaging Microscopy (FLIM) measurements into the dataset.
The end point is a fully automated processing workflow to detect subtle changes in organelle morphology. The research involves improvement of a FLIM protocol for speed and accuracy, optimize modulation and illumination parameters, develop noise reduction algorithms. By using active feedback loops for real-time measurement adjustment, the harvest of target patterns will be optimized using reinforcement learning or other control algorithms for digital microscopes.
Location: Groningen, The Netherlands
Supervisor: Prof. Dr. L.R.B. Schomaker (l.r.b.schomakerATrug.nl); Mr. J. Wehmeijer (jeroenATlambertinstruments.com)
Co-supervisors: Prof. Dr. I.J. van der Klei, University of Groningen & Prof. Dr. M. Fransen, KU Leuven
Planned secondments: KU Leuven, University of Groningen
Required subject specific skills and expertise: a master’s degree in computer science or artificial intelligence with specialisation in image processing and (deep) machine learning. Experience with biological imaging techniques is advantageous but not mandatory.
Application deadline: the application deadline has been extended until december 10th
Starting date: April 1, 2019