Biomedical Physics Group

Wed, 12 May 2021 00:00:00 +0000

Tue, 13 Apr 2021 00:00:00 +0000

Mon, 26 Oct 2020 00:00:00 +0000

CaosDB Max Planck Digital Library Project

Wed, 20 Nov 2019 00:00:00 +0000

CaosDB is an open source research data management system (RDMS) which was developed by the Research Group Biomedical Physics. This page informs about a project funded by the Max Planck Digital Library with the following aims:

Extend the research data management system CaosDB and prepare it for intergroup usage at the MPI for Dynamics and Self-Organization
Improve documentation of interfaces and clients and creation of tutorials
Enable more scientists internal and external to the Max Planck Society to benefit from the advantages of CaosDB

Project Information (German)

Titel: CaosDB – Dokumentation, erweiterte Softwareschnittstellen und Performance-Charakterisierung eines semantischen Forschungsdaten-Management-Systems für heterogene wissenschaftliche Umgebungen
Projektverantwortlicher: Stefan Luther
Projektdauer: 2 Jahre
Abstract: Deutsche Version / Englische Version

Abstract

English

Improving data management is one of the major challenges in all scientific fields. While there is a broadly accepted consensus about the aims and principles (e.g. FAIR data principles) there is a serious deficit of tools and established workflows tageted at their implementation. A special challenge is posed by the dynamic nature of scientific work rendering classical data management systems impractical. Often they are too specialized, not prepared for coninuous changes of data models or too generic to meet individual requirements.

The Open-Source software CaosDB is targeted at the integration and management of a high variety of complex data during the complete life-cycle, including data acquisition, data analysis, publishing and archival. Its unique semantic and dynamic data model does not only allow for powerful searches and reliable automatation, but also the flexible adaption of data structures by the end user.

The aim of the proposed project is to extend the research data management system CaosDB and prepare it for intergroup usage at the MPI for Dynamics and Self-Organization. This does not only include the technical implementation, installation and connection to existing data processing facilities, but also the training of scientific staff and data management representative. During the project, the documentation of interfaces and clients will be improved and tutorials will be created, enabling a higher number of scientists internal and external to the Max Planck Society to benefit from the advantages of CaosDB. Finally a performance evaluation will be conducted that paves the way to turn CaosDB into a generally accepted tool for high throughput data analysis.

Deutsch

Wissenschaftliches Datenmanagement steht heute mehr denn je im Fokus. Während man sich einerseits über die Ziele und Prizipien relativ einig ist (z.B. FAIR-Principles) so fehlt es in der Praxis häufig doch an Werkzeugen um diese umzusetzen. Mit seiner sehr dynamischen Natur ist der Wissenschaftsbetrieb eine große Herausforderung für klassische Datenbanken. Sie sind oft zu spezialisiert und nicht auf ständige Anpassungen von Datenstrukturen ausgelegt oder zu generisch um individuellen Bedürfnissen gerecht zu werden.

Die Open-Source-Software CaosDB bietet die Möglichkeit, vielfältigste Daten während ihres gesamten Lebenszyklus von der Aufnahme, über die Weiterverarbeitung und Veröffentlichung bis zu Archivierung abzubilden. Das semantische Datenmodell erlaubt dabei nicht nur mächtige Suchfunktionen und verlässliche Automatisierung der Datenverarbeitung, sondern auch das flexible Anpassen der Strukturen durch die Nutzer.

Im Rahmen dieses Projektes wird das Forschungsdatenmanagementsystem CaosDB für den arbeitsgruppenübergreifenden Einsatz am MPI für Dynamik und Selbstorganisation erweitert. Dazu gehören sowohl die technische Umsetzung der Installation und der Einbindung in die Datenverarbeitung als auch die Schulung der wissenschaftlichen NutzerInnen und verantwortlichen Datenmanagementbeauftragten. In diesem Zuge werden auch die Dokumentation der Schnittstellen und Clients verbessert und Tutorials erstellt, damit eine noch größere Zahl von WissenschaftlerInnen auch außerhalb der Max-Planck-Gesellschaft die Vorteile von CaosDB nutzen können. Zudem findet eine Performanceanalyse statt, um die Entwicklung von CaosDB zu einem geeigneten und allgemeinverfügbaren Werkzeug für Hochdurchsatzdatenanalyse voranzutreiben.

More resources on the MPDL Project

Link to MPDL project page (Max Planck internal only): https://www.soli.mpdl.mpg.de/de/vorschlaege_foerderung/gefoerderte_antraege/

More resources on CaosDB:

Official Webpage: https://caosdb.org
Article about CaosDB: https://www.mdpi.com/2306-5729/4/2/83/htm
Sourcecode, Wiki and Documentation: https://gitlab.com/caosdb
Slides from a recent talk
CaosDB page on this website

CaosDB

Mon, 14 Oct 2019 00:00:00 +0000

Overview

CaosDB is an open source research data management system (RDMS) which was developed by the Research Group Biomedical Physics. Within our research group CaosDB is used to manage all kinds of data including:

Experimental raw and processed data
Data from cardiac simulations
Data analysis results (plots, tables, …)
Scripts, tools, settings, parameter files which were used to carry out processing, simulations and data analysis
Publications like posters, articles, presentation files

The links between the above mentioned data are stored within CaosDB and allow for traceability of data during the whole data life cycle.

Key Features

CaosDB has several unqiue key features which make it specifically useful for scientific data management:

Automatic data crawling and indexing:
- No manual insertion of data required (although additionally possible via graphical web user interface)
- Existing structures and workflows can be reused: No data has to be moved to make use of CaosDB.
- Coexistence of CaosDB with other workflows: If necessary, all data can be accessed without CaosDB using the file system.
Flexible semantic data models allow for continuous adaption to changing scientific workflows
Intuitive Query Language (CQL) which can be used interactively or from scripts

More resources on CaosDB:

Official Webpage: https://caosdb.org
Article about CaosDB: https://www.mdpi.com/2306-5729/4/2/83/htm
Sourcecode, Wiki and Documentation: https://gitlab.com/caosdb
Slides from a recent talk
Slides from another talk about the CaosDB crawler
Project page of the MPDL Project

MultiRecorder

Mon, 01 Jan 0001 00:00:00 +0000

MultiRecorder is a software developed in our group for simultaneous recording from multiple scientific cameras. Key features are it’s ease of use, flexible configurability, live analysis, and remote control.

This makes MultiRecorder the perfect tool for multi-camera panoramic optical mapping of cardiac electrophysiology. But of course, it can be also used in other experiments or with just a single camera.

Feature list

General

Simultaneous recording from multiple cameras.
Support of many different camera models via plugins: see table below.
RAW data storage.

Display and Analysis

Live display of camera image with histogram, contrast adjustment and false-color display.
Select pixel region to extract time series (trace).
Traces window, showing time series of all cameras (on top of each other or vertically stacked).
Automatic vertical scaling of traces.
Sectrogram window, showing frequency spectrum computed from traces.
Filters for:
- Bandpass
- Frame Difference
- Ratiometry

Recording

Meta data (date, time, number of recorded frames, camera settings, etc.) of each recording stored in a machine-readable XML file.
Possibility to write image data from multiple cameras into a single file to reduce disk load and enhance data throughput.

Camera settings

Individual setting of camera parameters for each camera.
Saving and loading of configurations.
Multiple named configurations per camera.
Possibility to select the same or different configurations for preview and recording. This allows

Automation

Remote control via network TCP/IP connection from another program or computer, allowing integration in automated workflow and data management systems.
Built-in automated recording and software-triggering.

Supported cameras

Various cameras from different manufacturers are supported via plugins.

Manufacturer	Library	Tested with these models
Allied Vision	VimbaC	Pike F032B ASG16
IDS Imaging Developmet Systems	IDS Software Suite	UI-3060CP-M-GL Rev.2
Photometrics	PVCAM	Cascade 128+, Cascade II 512, Evolve 128
QImaging	QCam	RETIGA EXi, QICAM FAST
General DCAM cameras (Firewire, IEEE 1394)	libdc1394
Brainvision Inc.	MiCAM Ultima 3 System SDK	MiCAM ULTIMA-L (in development)

Note: Not all possible camera settings or features may be implemented by the respective plugins.

Planned improvements for future releases

Live synchronization of settings of multiple open cameras: Possibility to specify that a parameter should follow the setting of another selectable configuration (from the same or other camera), for each parameter individually.
Modularized filter system for live image analysis (e.g. to visualize exitation waves in phase contrast recordings of cell cultures of cardiomyocytes).
Multiple traces per camera.
Documentation, code cleanup and release as open-source software.

Community

MultiRecorder is available on inquiry.

An open source release is planned for the near future.

We are looking for developers. Knowledge of C++ and Qt required.