n-D Biomedical Imaging : from Instrument to Inference Journée inter-instituts d’Alembert / Farman 18 mars 2010

B. Chalmond

• ENSC-CMLA - Y. Yu, A. Trouvé, B. Chalmond • Institut Pasteur - S. Shorte, O. Renaud, (Imagopole/PFID)

n-D microscopy and quantification A High-resolution 3D Imaging System for Non-adherent Living Cells ENSC CMLA – Pasteur PFID -

European project STREP: AUTOMATION ACI NIM : FLUTOMY CNRS grant ANR Emergence, Inserm Transfert : NIN-ALL cell viewer

Confocal Z-stack

-International Patent,CNRS/Pasteur 2008

3D confocal imaging from microrotation images series

Main purpose

Main purpose:  to deal with individual non adherent living cells  to alleviate the anisotropy resolution and axial aberration

x-z image of a beads (0.17 μm). The resolution perpendicular to the focal plane is at least two times lower than the resolution within the focal plane.

(a) z-stack view,

(b) Micro-rotation reconstruction

Cultured SW13/20 living cells (human tumour cell lines)

Instrumentation for cell capture and manipulation

On the right: Electricfield (red values correspond to high intensity) and combined particle guiding force distribution (arrows) in the central plane between the electrode planes.

Scheme of dielectrophoretic octode field cage with trapped cell

DFC cage____

One large rotation

A continuous rotation Two types of rotation____

Multi-Z-stacks protocol Multiple axial through-stack sampling of fluorescent cell using dieletrophoretic field micro-rotation. (A) A schematic representation of the imaging target paradigm and acquisition protocol. The object is shown in two positions representing translation by DFC-induced perpendicular rotation wherein a through-stack confocal image series was acquired (gray volume divided into zaxial sections) for each position yielding two z-stacks (“zstack1” and “z-stack2”). Note: black X, Y, Z axes represent the geometry of the optical imaging system, whereas the red, blue, yellow x, y, z axes represent the assigned geometry of the cell relative to the first z-stack view (the origin). z-stack 1 and z-stack 2 were then combined to generate a fusion-deconvolution stack. (B, C) Confocal fluorescence z-stacks recorded from a single living human cell maintained and rotated in suspension by DFC manipulation. Two z-stacks were acquired from the same cell trapped in suspension with an estimated angle of 105.42 degree (compare plane XY of z-stack 1 in B with plane XY of z-stack 2 in C). Note z-stack 2 (C) has been registered to z-stack 1 (B). Axial reconstructions of each z-stack are annotated in the top right of each image (XZ and ZY). The acquisition time to record z-stack 1 and 2 was 2.8 s (96 frames with an exposure time of 30 ms). The time delay for the rotation was 7 s. Through-stack acquisition used the Andor spinning-disk confocal imaging microscope equipped with 63× water (1.2NA) objective. The z-step acquisition step was 100 nm and the scale bar represents 5 μm.

3D reconstruction from two z-stacks

Fusion-deconvolution 3-D volume reconstruction. (A–A’’) Results of the fusion-deconvolution of z-stack 1 (from Fig. 4B) and z-stack 2 (from Fig. 4C). These data can be compared for quality improvement with a conventional deconvolution of z-stack 1 (B-B’’). (B’, B’’) z-stacks 1 are respectively orthogonal views XZ, and ZY that have been extrapolated directly from XY view (B). Note how the fusion-deconvolution reconstruction improve the z-resolution, comparing A’ and A’’ with B’ and B’’, and therefore corrects for optical aberration. Scale bar in (A) represents 5 μm.

Biotechnology Journal, 2008, 3.

Confocal micro-rotation acquisition

Main difficulty : - the rotation unstability - the slice positions are unknown

100 - 300 CMR images per full rotation

In the case where only the CMR series is available :

The method for 3D reconstruction we reported previously suffered a significant burden of computation.

Inverse Problems, 2008, 24.

Bi-protocol : z-stack + CMR series

Journal of Microscopy, 2010

Reconstruction procedure

Journal of Microscopy, 2009, 233.

Comments

We deal with a multiple-slices-to-volume registration. It means that the registration treats the CMR slice alignments as a single set and not as a sequence of independent alignments. Let us again emphasize that z-stack volume and CMR volume are quite different . In contrast to CMR acquisition, z-stack data suffers from several serious drawbacks :    

axial aberration including spherical aberration, poor axial resolution, spatially z-varying point spread function, photobleaching.

z-stack

CMR reconstruction

Artificial data

Acquisition

The experiments shown here were performed on a sequence of real micro-rotated images. Cultured SW13/20 living cells (human tumour cell lines) tagged nuclear targeted green fluorescent protein (lamin-A-GFP, a kind gift of Christopher Hutchinson) were suspended in a DFC-3 chip (Evotec Technologies/Perkin Elmer group, Germany) controlled by a Cytocon400TM’s 4-phase high frequency generator (Evotec Technologies). Individual cells were rotated around the x–y axis and imaged using an Andor Revolution XD spinning disk confocal system equipped with an EM-CCD DV885 camera (Andor Technology, Belfast Northern Ireland)mounted on an inverted microscope (Axiovert 200M, Carl Zeiss, Germany). The microscope is equipped with a 63× water immersion objective with a numerical aperture (NA) 1.2. Fluorescence acquisition used laser light excitation 488 nm and emission band-pass filter 500–550 nm.

Furthermore, our spinning disk confocal microscope is equipped with an objective piezodrive. So, in addition to the micro-rotation sequence, a so-called axial “through-stack” image series (or “z-stack”) is recorded from the target samples immobilized in suspension, (see Fig. 4). The piezo step in through-stack axial imaging was 100 nm and xy resolution was 127 nm. More details about z-stack acquisition can be found in (Renaud et al., 2008). To measure the point spread function, 3D image stacks are acquired from subresolution beads (0.17 μm, Molecular Probes) suspended in the same medium used for live-cell imaging, and using the same microscope configuration. The calculated axial resolution is 591 nm (microscope resolution calculator1) and the axial sampling interval is 100 nm. Image data from five to eight independent measurements were averaged and the PSF is calculated using Huygens Pro software (Scientific Volume Imaging the Netherlands). The 3D image is given from a zstack of this bead according to the protocol used for acquiring cell images.