Microscopy Conference 2023

Prof.

Session chairs

Modern microscopes are capable of quickly producing terabytes of multidimensional and multimodal data in form of (correlated) images, spectra, diffraction patterns. While managing and (online) processing such large amounts of data may require some investment in computer hard- and software, the even bigger challenge lies in the effective utilization of such data, often in an international team. Ultimately, the paradigm of FAIR (findable, accessible, interoperable, and re-purposable) data aims at data re-utilization by others, potentially with completely different research questions in mind. 
This session welcomes contributions describing solutions that enable handling and processing of big data in microscopy, but also solutions that make microscopy data FAIR.  An important aspect here, is the (automated) annotation of the experimental data by relevant metadata that is both human and machine readable. In addition, this session shall provide room for the demonstration of examples where FAIR research data management helps to lower the threshold for sharing data with collaborators and the scientific community and as such enabling synergistic combining of data from various sources or techniques. 


Prof. Dr.

Putting your data into the void

Abstract text (incl. figure legends and references) With rapid advances in detector technology and computing resources, modern microscopes produce massive amounts of data. Additionally, advanced sample preparation techniques often involve various steps utilizing multiple devices each generating additional data. Therefore, data management should be considered a crucial part of today's research infrastructure. This lead to the publication of guiding principles for scientific data-management describing four fundamental cornerstones: Findability, Accessibility, Interoperability and Re-usability (FAIR) [1]. To deal with this, we created a system capable of handling most of the usual data management tasks in an automatic fashion without getting in the way. This is implemented by a file storage infrastructure automatically collecting data from various experimental devices including electron microscopes, optical microscopes, sample preparation devices and mechanical testing systems. Raw research data is considered immutable and is therefore stored read-only on a scalable distributed storage system [2]. The data can be accessed via a front-end server by a secure connection using asymmetric cryptography. Metadata is handled via an electronic lab book system [3]. This system is accessible via standard web browsers and is used for documenting experimental steps and persons, devices and samples involved in carrying them out. The lab book system is also used for keeping a database of samples. Currently we are testing an automatic sample tracking workflow based on QR-codes and a distributed database populated from the lab book entries. A complete research data management system also has to cope with data processing. To achieve a reusable and well-documented data processing and analysis workflow, we use GitLab [4] to track development of our analysis tools and manuscripts. The GitLab server also acts as an issue tracker for administration of all the involved hard- and software assets. To setup a future-proof, independent system and to avoid possible vendor lock-in scenarios, care was taken to only use free and open-source software [5]. In its current state the system is used internally in our work-group. A workflow for publishing data for external use is planned via the Repository for Permanent Hosting, Archiving and Indexing of Digital Resources and Assets (PHAIDRA) at the University of Vienna [6]. References [1] M. Wilkinson, M. Dumontier, I. Aalbersberg, et al., "The FAIR Guiding Principles for scientific data management and stewardship", Sci Data 3, 160018, 2016, doi:10.1038/sdata.2016.18 [2] S. A. Weil, S. A. Brandt, E. L. Miller, D. D. Long, C. Maltzahn, "Ceph: a scalable, high-performance distributed file system", In Proceedings of the 7th symposium on Operating systems design and implementation (OSDI '06). USENIX Association, USA, 307–320, 2006 [3] N. Carp, A. Minges, M. Piel, "eLabFTW: An open source laboratory notebook for research labs", J. Open Source Softw., 2(12), 146, 2017 [4] P. Choudhury, K.Crowston, L. Dahlander, et al., "GitLab: work where you want, when you want", J. Org. Design 9, 23, 2020, doi:10.1186/s41469-020-00087-8 [5] R. M. Stallman, "GNU's Bulletin, Volume 1 Number 1". https://gnu.org/bulletins/bull1.txt. p. 8. (accessed 2022-07-28) [6] University of Vienna, https://phaidra.univie.ac.at/ (accessed 2022-07-28).

Workshop 1: Data management

Invited talk

A community-driven glossary for data interoperability in the electron microscopies

Abstract text (incl. figure legends and references) Research data has to fulfil a number of requirements for it to be fully exploitable - today and in the future. The FAIR principles [1] – findable, accessible, interoperable, and reusable – provide a conceptual framework to achieve this, but the pathway to implementation is often unclear. Here we focus on the challenge of implementing the &quot;I&quot; in FAIR. Interoperability is the potential for two independent agents to work on the same data in a coordinated fashion. It is especially relevant for researchers working in interdisciplinary fields, such as electron microscopy, where terminology in adjacent scientific disciplines is unknown, or where identical terms are used with differing meaning, for example, consider a mathematical or biological cell. The use and adoption of established vocabularies and semantic standards can provide solutions – however these may be infeasible to implement, non-existent, or scientifically unsuitable. This is why a number of initiatives are developing semantic artefacts that aim to describe experimental equipment, workflows, and analysis procedures in electron and ion microscopies. Harmonisation of these efforts to ensure interoperability across the community is required. To support the long-term semantic interoperability the Helmholtz Metadata Collaboration (HMC) is coordinating a community-wide effort, including over 45 scientists from more than 22 institutions across Switzerland, Austria, and Germany, and including representatives of several NFDI consortia [2], to create a joint resource to harmonize semantics in the field of electron and ion microscopies. In our presentation we will report on this community effort - the EM glossary group [3]. We will describe how we achieve consensus on terms commonly used in electron and ion microscopies via a remote, collaborative workflow based on the platform GitLab. A text-mining approach has been used to evaluate EM image metadata and prioritise terms which are widely used in the community. We will show results, including some of our harmonized definitions, and provide details of our implementation strategy. The developed resource will provide harmonized and machine actionable semantics to support specific development efforts such as metadata schemas or ontologies in their respective fields. This will benefit experimental scientists by providing clarity when annotating and re-using research data; and by facilitating communication when conducting interdisciplinary studies. Do you work in electron and/or ion microscopy or related data analysis? Follow us on twitter @helmholtz_hmc or tweet #EMGlossary. You want to get involved in the development and/or join our meetings? Send an email to hmc@fz-juelich.de or hmc-matter@helmholtz-berlin.de 1. Wilkinson, M.D. et al. Scientific Data. http://dx.doi.org/10.1038/sdata.2016.18 (2016) 2. NFDI: Nationale Forschungsdateninfrastruktur. https://www.nfdi.de/ 3. EM Glossary GitLab repository: https://codebase.helmholtz.cloud/em_glossary/

LS 4: Image analysis of large data sets

Dr.-Ing.

On cloud computing tools for the electron microscopy community and interoperable computational-geometry-based modeling of microstructures

Abstract text (incl. figure legends and references) Software is a fundamental tool which is used in virtually all workflows where data of materials characterization experiments with electron microscopes gets transformed into insight and knowledge. The last years have seen an increased interest, awareness, and funding to support the electron microscopy community on a common goal: Namely, data and metadata of electron microscopy experiments should become easier findable, more transparently accessible, and fundamentally more interoperable; so that data can be processed with different tools to improve the reusability and thus value of electron microscopy data [1]. Clear vocabulary, transparently detailed data schemes and software tools are needed to achieve this aim. This is the essential idea behind the FAIR research data stewardship principles [2] and the key reason behind established national funding schemes like the German National Research Data Infrastructure [3], Platform Material Digital [4], and international activities like EOSC [5], and e-DREAM [6]. A large number of workflows are currently developed by scientists using specific software tools locally and running analyses on own computers and high-performance computing resources. Manual data transfer and documentation is still a frequent task. Usually, many file formats and data representations have to be mastered. However, inconsistencies can come with having multiple representations of similar or sometimes even the same type of data. Reference frame conventions or assumptions used are examples which have to be carefully documented and communicated across tools to ensure a correct and valid interpretation of data. For this task, often recent efforts in cloud computing, documenting data via electronic lab notebooks (ELNs), and using research data management systems (RDMS) are expected to support scientists. Clearly, these tools can be useful in that they offer data organization, analysis, and publication services. However, using all these tools without changing culture and and rigorous data documentation, does not necessarily leave us with FAIR data. We will show how we are addressing these topics within the FAIRmat [7] project using NOMAD OASIS as the RDMS [8]. Specifically, this talk pulls focus on two topics relevant for electron microscopy: First, we will show the services which NOMAD offers, usage of ELNs and cloud computing services as examples. We will briefly touch upon the data model that is behind this work [7, 8] and the value of working as a community on interoperable data models [9]. Second, we would like to discuss the specific topic of analyzing diffraction pattern, here with a focus on orientation microscopy. We suggest an approach which can describe the variety of representations of orientation microscopy data. We will show an initial implementation of such an interoperable description using ELNs and NOMAD that can also be used for description of geometrical data like interface network and crystallites. [1] M. Scheffler et al. Nature 604 (2022) 635 [2] M. Wilkinson et al. Scientific Data 3 (2016) 160018 [3] https://www.nfdi.de [4] https://www.materialdigital.de [5] https://eosc-portal.eu [6] https://e-dream-eu.org [7] https://fairmat-experimental.github.io/nexus-fairmat-proposal [8] https://nomad-lab.eu [9] https://codebase.helmholtz.cloud/em_glossary [10] https://arxiv.org/abs/2205.13510 The project is funded by the Deutsche Forschungsgemeinschaft (DFG, German Research Foundation) – project 460197019.

IM 3: SEM and FIB developments

Integration of multi modality image data in structural lung research

PD Dr. sc. hum.

Invited speaker

Electronic lab notebooks compiling big datasets for machine learning analysis

Abstract text (incl. figure legends and references) Electronic lab notebooks (ELNs) serve as means to gather analog metadata, e.g. experimental parameters, that would otherwise be hard to digitalize. Furthermore, they are a key prerequisite to a comprehensive documentation of research processes and the reuse of research data. However, different systems are often used within the same research institution or community, especially when covering a long, interdisciplinary process chain. The use of different systems - each addressing distinct requirements for discipline-specific needs - enables the availability of a broad functionality but results in challenges due to an often missing interoperability of the metadata. We are addressing this lack of interoperability for the two ELNs Herbie and Chemotion with an API-based data exchange. The envisaged reduction of boundaries between the disciplines of chemical synthesis and process engineering in membrane research will enable the creation of large and coherent data sets, including microscopic image data. Once data management platforms, e.g. ELNs, are in use, the semantic information therein can only be exploited when terms are clearly defined in a glossary. In order to make metadata truly interoperable, glossaries need to be transformed into ontologies, which structure knowledge based on formal logic, reduce ambiguity and offer machine-driven reasoning. We are establishing a glossary and ontology for membrane research, in parallel to a guide that describes the best practices for building these. In combination with existing efforts, e.g. the Helmholtz Electron Microscopy Glossary, our ontology will enable machine-driven interpretation of the linguistic data in our ELNs. Following the assembly of large datasets in ELNs, we address the ambiguous relationship between pore morphology and fabrication parameters of isoporous block copolymer membranes by machine learning methods. At first, the hexagonality was extracted from 1800 SEM images as a quality descriptor. Analyzing the impact of the fabrication parameters on this descriptor with a neural network revealed that the fraction of the main solvent and the molecular weight have the most pronounced effect on the hexagonal arrangement of pores in the membrane. In summary, using ELNs in the entire experimental workflow creates holistic, structured datasets that can easily highlight critical parameters when analyzed by machine learning methods. Moreover, enhancing the value of the metadata in ELNs by an ontology creates even more room for digital analyses.

Abstract talk

Automation of TEM alignment using python scripting

Abstract text (incl. figure legends and references) In recent years, image processing technology has been greatly improved and people in various fields have developed automated operations based on image processing technology. In the field of electron microscopy, automated functions and software have been developed that can be used in the fields of materials and biology [1]. Transmission electron microscopes (TEM) of JEOL can be operated using a scripting language with PyJEM, a JEOL made Python module (Python is a trademark or registered trademark of the Python Software Foundation). By using PyJEM, one can read the information on a TEM and send operational commands to it without a direct operation. In addition, the combination of lots of computation and image processing modules enables users to simplify the operation of TEM together with the acquisition of experimental data. Moreover users who are not familiar with microscope operation can perform TEM alignments with the same accuracy as an experienced user in a short time by combining useful functions in Python. In this report, we introduce examples of automated microscope alignments by PyJEM. [1] M. Schorb. I. Haberbosch. W.J. H. Hagen. Y. Schwab &amp; D. N. Mastronarde. Nature Methods 16 (2019). p. 471-477.

WS 1
Data management

Appointment

Session chairs

Description

Programme

WS 1Data management

Appointment

Session chairs

Description

Programme

WS 1
Data management