Case Studies

Advanced 4D tracking of meiotic spindle dynamic


Gunar Fabig, Müller-Reichert lab, Experimental Center, Medical Faculty Carl Gustav Carus, Technische Universität Dresden (TUD), Germany

Live-cell imaging of meiotic events in C. elegans

With its expertise on correlative light and electron microscopy (CLEM), the laboratory of Prof. Dr. Thomas Müller-Reichert at the TU Dresden is committed to study the details of meiotic cell divisions and the function of the spindle apparatus during this process using the model organism C. elegans. Whereas mitotic spindle dynamics have been studied in detail within C. elegans, the dynamics during male meiosis remain mainly elusive. Using the live animal, Gunar Fabig, PhD student at the Müller-
Reichert lab, conducted for the first time a quantitative analysis of spindle dynamics during male meiosis I and II in 4D. Combining the advantages of C. elegans (transparency, easy mounting, the possibility to fluorescently tag proteins, accessibility of the gonad) with spinning disc microscopy (fast image acquisition), Gunar was able to obtain 3D stacks of the full male gonad every 30 s for 45 to 60 min.

Tackling challenges in image analysis with arivis Vision4D

Analyzing the spindle pole dynamics of wild-type C. elegans in these data sets was quantitatively approached by measuring the spindle pole-to-pole distance over time. By doing so, the researchers were confronted with two major challenges. First, due to 4D imaging of live animals, every spindle was randomly located and orientated within the data set. Second, although mechanically immobilized, a slight movement of the roundworm could not be prevented. These conditions necessitated a demanding image analysis including 4D segmentation, 4D tracking and the possibility to correct tracking failures as a consequence of sample movements.Combing his imaging approach with arivis Vision4D, allowed Gunar to tackle the challenges, extract the desired information in an efficient manner and develop a standardized analysis pipeline. He specifically benefitted from the possibility to analyze randomly oriented spindles by calculating the center of mass of the spindle poles and to edit wrong tracks very easily with the track editor.
Using this interactive tool, individual tracks could be easily identified, selected and edited, which includes merging or splitting of tracks by a simple drag and drop function.

“The segment tracking enabled us to analyze the spindle elongation over time very efficiently. So far, it is the only working solution for us to  analyze our data to our full satisfaction.”

(Gunar Fabig)

Analysis pipeline using arivis Vision4D

Based on these advantages, Gunar Fabig created an analysis pipeline within arivis Vision4D to standardize the analysis of spindle elongation:

  • After import, a denoising filter was applied.
  • Individual meiotic events within the gonad were analyzed separately by creating 3D Regions of interest to export single spindles.
  • Within these ROIs, spindle poles marked by GFP were defined in 3D using an intensity threshold segmentation.
  • To exclude small, false positive segments, a filtering step based on volume size was applied.
  • The segmented spindles were tracked over time by using the arivis Vision4D tracking module including the track editor.
  • The analysis report was exported as Excel file including the center of mass coordinates of each segment over time.

Towards a big picture of the meiotic spindle

Based on this fast, quantitative analysis of spindle dynamics in wild-type C. elegans using arivis Vision4D, the Müller-Reichert lab is now able to further complete the big picture of the meiotic spindle and address outstanding questions. Altogether, this analysis will shed light onto the molecular mechanism, which allows the spindle apparatus to accurately segregate paired and unpaired chromosomes. Visit:

Fig. 1: 4D segmentation and 4D tracking of spindle dynamics; (left) 3D representation of male meiosis I in living C. elegans; spindle poles are marked by GFP fused to γ-tubulin, chromosomes are marked by mCherry fused to histone H2B. Segmented spindle poles are shown in green, chromosome are shown in red using the 4D Viewer of arivis Vision4D. Tracks of segmented spindle poles over time are depicted as lines. The left track is selected and can be easily identified using the track editor (right), highlighted in blue. Within the track editor, individual tracks are shown on the left, time frames on the top. Every segment is visualized as a small circle. A series of connected segments that form a track is visualized by a horizontal line connecting the segment circles. Merging or splitting of tracks can be easily done by clicking on segments and dragging them to their desired position.

Source Data set specifications
Microscope: Olympus IX83, invers; Andor IQ 3.2; equipped with Yokowaga CSU-X1, Andor iXON Ultra 897 EMCCD detection camera and Olympus U Plan S Apo 60x / 1.2 W objective; Data size: 1 GB to 3 GB; Time series: stack acquisition every 30 s for 45 to 60 min; Picture size: 512 x 512 px; Voxel size: 0.222 µm x 0.222 µm x 0.33 µm; z-depth: ~60 planes, 19,8 µm;
arivis Software Package
arivis Vision4D: Base Module, Analysis and Tracking Modules




The migration of interneurons in the developing mouse brain


Tibor Harkany Lab Department of Molecular Neuroscience, Center for Brain  Research, Medical University of Vienna - Daniela Calvigioni, Zoltán Máté, János Fuzik, Fatima Girach and colleague

Imaging mouse brains using Light-Sheet Microscopy

The laboratory of Univ. Prof. Dr. Tibor Harkany at the Center of Brain Research in Vienna is interested in the diversification of neurons and their integration into neuronal networks during development. To add knowledge to the diversity of interneurons, Daniela Calvigioni and  colleagues analyzed a subtype of GABAergic interneurons producing the neuropeptide cholecysteokinin (CCK). Due to difficulties in histochemical approaches, the migration of CCK interneurons and their population of the cerebral cortex at prenatal stages is poorly understood. Therefore, the researchers developed a novel transgenic mouse line marking CCK interneurons in vivo. This opened the possibility to localize this interneuron subtype for the first time within the intact brain structure and to analyze its migratory behavior. To do so, dissected brains of different embryonic time-points were fixed, cleared (CUBIC clear-ing) and imaged on a ZEISS Lightsheet Z.1 microscope using tile scanning to cover the size of the entire brain.

arivis Vision4D as prerequisite for stitching and visualizing the mouse brain

After image acquisition, the individual z-stacks had to be stitched to generate a 3D image of the mouse brain. However, the researchers were confronted with the enormous data set of ~100 GB for one channel only. This difficulty was only solved by arivis Vision4D. Fatima greatly benefitted from its ImageCore technology, which allows the fluent processing of imaging data with unlimited size on standard hardware. “arivis Vision4D can handle big data, it can easily open hundreds of GB files, stitch them and create your 3D image without slowing your system down.” Having this essential tool in her hands, Fatima developed a standardized image processing pipeline:

  • After import, individual z-stacks were aligned and stitched  using the Tile Sorter
  • Resulting mosaics were visualized in 3D using the 4D viewer
  • Movies were generated  using the storyboard feature

Handling large data sets to understand interneuron heterogeneity

This study benefitted from arivis  Vision4D as the only software with the ability to easily stitch, visualize and share large data sets. Therefore, only the combination of Light-Sheet Microcopy and arivis Vision4D allowed the researchers to obtain 3D information of the entire brain and share this via movies in a very demonstrative way. Based on this, a classical tangential migratory route of CCK interneurons while populating the cerebral cortex was ob-served. Overall, this study identified several characteristics of prenatal CCK interneurons and integrates important information for unravelling the diversity of interneuron functionality people/persons-details/id/301 cercor/article/27/4/2453/ 3056356/Functional- Differentiation-of-Cholecystokinin

Fig 1: Stitching of individual tiles using the Tile Sorter; shown are 35 individual z-stacks covering a dual labelled mouse brain (CCKBAC/DsRed::GAD67gfp/+) at embryonic day 16.5; GA-BAergic neurons are marked by GFP (green) due to expression of GAD67; CCK expression is marked by DsRed (red). Tile 4 is selected and highlighted by a white box. For aligning, the Tile sorter provides different methods including Grid, Manual and Alignments using advanced algorithms. In this case, the z-stack tiles were sorted using the Grid mode; Pixel overlap was set to 10% as during imaging. Several planes were checked to assure correct alignment. Due to stability during image acquisition no further adjustments as algorith-ms were used. Scale options and transparency settings (see selected Tile 4) help to assure correct alignment.

arivis Big data imaging solutions

rivis AG, headquartered in Munich, Germany, is a market leading software company focused on the life sciences industry. arivis AG provides imaging solutions in multi-dimensional microscopy for datasets of basical-ly unlimited file size based on the in-house developed  ImageCore technology. With our desktop software  arivis Vision4D, scientists are empowered to work with terabyte sized images fast and efficiently on ordinary workstations and laptops. Additional benefit to usability and performance is the possibility to apply color mapping,rendering methods or quantification intuitively with imme-diate feedback and preview of the corresponding results. This potential can be scaled up with arivis WebView, a server-based image analysis framework that allows to access, display and analyze large image data in a standard web browser. With the world’s first and only virtual reality visualization system for real microscopy images, arivis InViewR allows scientists to gain all- dimensional insights by fully immersing into the data.


Source Data set specifications
Microscope: Lightsheet Z.1 microscope (Zeiss), ×5 (EC Plan Neofluar 5×/0.16) detection objective, ×5/0.1 illumination optics, PCO edge sCMOS camera; Imaging settings: ×0.7 zoom, laser power of 20v%, exposure times of 200 ms, tile scan overlap 10%; Data size: 100-300 GB; Picture size: up to 48 tiles, each 1200 px x 1200 px; Voxel size: 1,3 µm x 1,3 µm x 3,5 µm; z-depth: up to 1500 planes
arivis Software Package
arivis Vision4D Base Package