MOTILITY-FLOW AND GROWTH CONE NAVIGATION ANALYSIS DURING IN-
VITRO NEURAL DEVELOPMENT BY LONG-TERM BRIGHT-FIELD IMAGING Maya
Aviv and Prof. Zeev Zalevsky, Faculty of Engineering, Bar-Ilan
University M. Pesce, S. Tilve, E. Chieregatti and Dr. F. Difato,
Istituto Italiano di Tecnologia, Department of Neuroscience and
Brain Technologies, Genova, Italy Feb 2014 J. Biomed. Opt. 18 (11),
111415 (September 20, 2013)
Slide 2
Agenda Motivation Background Incubator-Imaging system Image
enhancement and processing results
Slide 3
Motivation Investigate motility flow and grown cone navigation
during early stage of neural development in order to learn about
the neurons growth mechanism Challenge: Long term imaging avoid
phototoxication, pay with low contrast
Slide 4
Neural structure Soma central part Dendrites - cellular
extensions with many branches Spine - a small part from a neuron's
dendrite that receives input Axon - is a finer, cable-like
projection. The axon carries nerve signals away from the soma and
back. Neurite refers to any projection from the cell body of a
neuron, when speaking of immature or developing neurons.
Slide 5
Neural Wave Growth cones are the main motile structure located
at the tips of the neurite. Image of a fluorescently labeled growth
cone extending from an axon Dr. Difato Francesco,
Photonic-Neurosurgery lab, Istituto Italiano di Tecnologia Some
neurons exhibit periodic recurring growth cone-like structures,
referred as "waves followed by growth bursts.
Slide 6
Incubator-Imaging System The whole micro-incubator
Slide 7
Image Enhancement and Processing Our goal is to develop an
image enhancement technique, based of time dependence morphology
techniques in order to monitor and measure the growing mechanism
over time and overcome poor imaging conditions: -~500 images per
movie -Low contrast -Non uniform illumination (space and time)
-Minor movements of the system (mechanical and biological)
Slide 8
Time Dependence Morphology The goal is to identify the
"significant" change, at a given a set of images of the same scene,
taken at different times The method is to compare each image to the
previous ones. A key issue is to deliver application (task)
specific differential morphology. Since finding the change mask is
usually the first step into understanding the change, segmentation
and classifying changes usually requires particular treatment.
Slide 9
Define Past Separate between constants (or slow changes) and
quick changes (Time derivative) by derivative with average set of
past images
Slide 10
(reminder - edge and open operators) Edge detection - an image
processing technique for finding the boundaries of objects within
images. It works by detecting discontinuities in brightness. Edge
detection is used for image segmentation. Open (morphology) - the
dilation of the erosion of a set A by a structuring element B:
Slide 11
Mark Significant Change mm
Slide 12
Output 94 images Time gap = 3 min ~5 hours
Slide 13
Results Actin Waves Bar is 10m Numbers indicate minutes
Slide 14
Results Actin Waves Actin wave velocity 30.5[m/min], appearing
with time gap of 395min
Slide 15
Results Tip Activity Bar is 10m Numbers indicate minutes
Slide 16
Results Tip Activity
Slide 17
Results Soma Activity 1 2 3 4 5 6 Soma area is divided into
sector. Sector activity is presented in time (temporal) and
spectrogram (45min)
Slide 18
Results Soma Sctivity PCA images shows short and long neurites
are similar in time and tempo, while undeveloped ones and growth
cones are different.
Slide 19
Summary Experimental system that represents a simple and non-
invasive approach to study neuronal growth Special image processing
algorithms were adapted Detection of very small and slowly moving
spatial changes, and to inspect low contrast image features
characteristic of motion and dynamics of a living cell in a long
time frame.