Alexander Ziegler on digital data for morphological studies

Posted by Biome on 19th May 2014 - 0 Comments


Earthworms hold a key position in the ecosystem as ecological engineers, sculpting the soil to aid hydration and oxygenation, which has made them a focus of research. Studies into their anatomy have traditionally been based on conventional dissection methods, however the advent of advanced imaging techniques, particularly non-invasive methods such as micro-computed tomography (microCT), have led to more detailed insights. In a GigaScience data note Alexander Ziegler from Ziegler Biosolutions, Germany, and colleagues from Harvard University, USA, present a vast resource of high resolution images, videos and interactive models of earthworm anatomy generated using microCT with soft tissue staining – this substantial dataset is available at GigaDB. Here Ziegler discusses the impact of the digital data revolution on morphological studies, the benefits of non-invasive imaging techniques, and also shares his top tips for scientists getting to grips with advanced imaging.

 

A 2 cm long specimen of the gastropod Cerithium caeruleum analyzed by microCT; 2D slice (top), 3D reconstruction (bottom). Image source: Alexander Ziegler, Ziegler Biosolutions

How did you first come across microCT?

During my PhD studies on comparative sea urchin anatomy, I first employed another non-invasive imaging technique, magnetic resonance imaging (MRI). Although this modality is well suited to visualize soft tissue structures, hard parts can only insufficiently be analyzed. Therefore, I asked a colleague to scan four sea urchins using microCT, an X-ray-based imaging technique. I was immediately struck by the speed of data acquisition and the high isotropic voxel resolution attainable. These first four scans were made available as series of images on DigiMorph. Since then, I have gathered microCT scans of almost two-hundred different sea urchin species (Zoosymposia. 2012, 7, 53–70) for my comparative analyses of mineralized structures. Lately, my colleagues and I have also employed soft tissue staining protocols (PLoS ONE. 2014, 9, 5:e96617) in combination with microCT to analyze museum and freshly fixed specimens.

 

What do you think has been the biggest benefit of non-invasive imaging techniques such as microCT or MRI for morphology?

Many of these imaging techniques have actually been around for a number of years or even decades, but large parts of the zoological community have, for various reasons, been hesitant to embrace them. However, I believe that MRI or microCT, among other imaging modalities, constitute excellent tools for one of the prerequisites for hypothesis-driven science, i.e. exploration. Natural scientists depend on observations of the world surrounding us to formulate their questions – interactively exploring the digital dataset of a scanned specimen in real-time on your computer provides an exciting opportunity for scientific discovery. In addition, non-invasive imaging techniques permit high-throughput studies of anatomical structures, thereby providing the basis for the correlation of geno- and phenotype on a large scale. I am therefore convinced that an increased use of digital morphological data is bound to trigger a new wave of scientific discovery.

 

In a 2010 commentary in Biology Direct, you stated that the digital data revolution that had transformed other biological disciplines, such as genomics, phylogenetics, and structural biology, had not yet found acceptance in morphology. Do you think much progress has been made since then?

Indeed! As the online repository GigaDB demonstrates, the previously seemingly insurmountable technical hurdles for the large-scale deposition of raw and derived imaging data have now been overcome. However, there remains a lack of coherent data management policies for digital morphological data. Furthermore, publishers, funding agencies, and professional societies have not yet agreed on a common approach to enforced data deposition prior to the publication of scientific findings, although this is standard practice for genomic or structural biological studies.

 

In your study in GigaScience, you reconstructed 2D projection images into a 3D volumetric image stack using the software NRecon. How easy it was to use this software to produce the final product and how accessible is this software?

Virtual sections through a stained earthworm obtained using microCT. Image source: Museum of Comparative Zoology, Harvard University. Photographer: Alexander Ziegler

Virtual sections through a stained earthworm obtained using microCT. Image source: Museum of Comparative Zoology, Harvard University. Photographer: Alexander Ziegler

Ease of data acquisition is one of the big advantages of non-invasive imaging techniques. Although the reconstruction process for data gathered during a scan varies depending on the imaging technique employed, all commercial scanners come equipped with the necessary software. In the case of the microCT experiments, this simply meant following the preset image reconstruction workflow outlined in the scanner manual. The difficulties usually start following data acquisition, in particular when large amounts of image data need to be rapidly analyzed or visualized in 2D and 3D, or have to be prepared for data deposition. These steps can sometimes be characterized by a steep learning curve, but are certainly not insurmountable for anyone interested in digital data analysis.

 

Any top tips for scientists wishing to experiment with similar imaging techniques?

Like any other technique used in science, non-invasive imaging techniques constitute tools. As such, they need to be chosen based on the scientific question being asked. There is certainly not a single technique that would be able to cover all possible imaging needs. However, scientists can now choose from a wide spectrum of non-destructive imaging tools that allow 3D scanning of specimens ranging from the millimetre to metre scale. In addition, new technical approaches such as episcopic imaging or wide-field scanning electron microscopy demonstrate that light and electron microscopy, respectively, offer a multitude of new opportunities for digital specimen analysis.

 

A 20 cm long museum specimen of the cephalopod Bathypolypus arcticus analyzed in a human MRI scanner. 2D slice (top), 3D reconstruction (bottom). Image source: Alexander Ziegler, Ziegler Biosystems

A 20 cm long museum specimen of the cephalopod Bathypolypus arcticus analyzed in a human MRI scanner. 2D slice (top), 3D reconstruction (bottom). Image source: Alexander Ziegler, Ziegler Biosystems

What do you think is the biggest hurdle for scientists in using advanced imaging techniques?

I presume that many scientists will, apart from methodological or logistic complications, shy away from using non-invasive imaging techniques due to fear of high usage fees. Indeed, depending on country and institution, the costs levied for access to MRI or microCT systems can be prohibitive and well out of reach for scientists wishing to scan multiple, let alone hundreds or thousands of specimens. Nonetheless, the total number of scanners around the world is rising at a steady pace, ultimately bringing down the cost of an individual scan. In addition, the recent investments in human diagnostics and industrial quality control will directly or indirectly result in easier access to non-invasive imaging equipment also for zoologists and palaeontologists. This said, there is currently a surprising reluctance, at least in my country, to fund high-throughput non-invasive imaging studies in morphology.

 

In your recent paper in PLoS ONE, you chose to publish the data note in GigaScience, specifically highlighting your data. Traditionally, the data behind an article has played a fairly silent role. How do you see new imaging techniques changing that?

All of the imaging techniques mentioned above are characterized by the production of large amounts of digital data right from the onset of an experiment. This is in stark contrast to conventional approaches in morphology, where digital data have always constituted a form of secondary, derived data (e.g. a digital photograph of a dissected specimen). However, to permit efficient data mining and for reasons of data transparency, access to raw data is essential. I believe that digital data from non-invasive imaging techniques will profoundly increase the importance of as well as the demand for primary data in morphological studies. However, only the future can tell whether the deposition of data in a database associated with a journal such as GigaScience or in a repository funded by a professional society or national funding agency will constitute the gold standard.

 

More about the author(s)

Alexander Ziegler, researcher, Ziegler Biosolutions, Germany.

Alexander Ziegler, researcher, Ziegler Biosolutions, Germany.

Alexander Ziegler is a researcher based at Ziegler Biosolutions, Germany. He has an interest in zoology, imaging, micro-computed tomography, magnetic resonance imaging, evolution, anatomy and morphology. Ziegler was previously based at Freie Universität Berlin, Germany, the Charité-Universitätsmedizin Berlin, Germany, and at Harvard University, USA.

Data note

A dataset comprising four micro-computed tomography scans of freshly fixed and museum earthworm specimens

Lenihan J, Kvist S, Fernández R, Giribet G and Ziegler A
GigaScience 2014, 3:6

Go to article >>