Managing and Processing Images from X-Ray and Neutron Facilities – University of Copenhagen

CoNext > Calendar > Managing and Processin...

Managing and Processing Images from X-Ray and Neutron Facilities

HCØ, Aud 9. 13-17, November 16th, 2017


13:00 Welcome

13:05 Software for Science - Hans Fanghor, XFEL
The talk outlines some of the challenges of the field of computational science which includes data processing and analysis and is increasingly regarded as a third pillar of research that complements experimental and theoretical work. The issues include the reproducibility of computational science, the lack of appropriate computational training for researchers at all career stages, the required high flexibility of research codes, and the existing performance metrics for researchers and research institutions.
I attempt to provide an overview of some emerging computational and workflow tools as well as software engineering techniques that can be used to start to address these challenges.

13:50 Artificial Intelligence and Big Data for Image Analysis: Segmenting and Characterizing Complex Samples at Scale - Kevin Mader, 4Quant
The diagnosis and treatment of lung cancer has been drastically improved by new imaging
methods which generate over 1 billion pixels per patient where single spots can drastically
influence the diagnosis and treatment. For physicians this means a long time must be spent carefully reading images. 4Quant (an ETH Spinoff) together with the University Hospital Basel have demonstrated the potential to radically reduce the reading time without sacrificing quality by using Big Data and Deep Learning approaches. We present the work we have done towards a computer aided staging of Non-Small Cell Lung Cancer (NSCLC).

14:35 Coffee

15:00 Novel Neutron Imaging Techniques and Processing Requirements at ESSMarkus Strobl, PSI
While neutron imaging has proven its vast potential for scientific and industrial applications at continuous sources, neutron imaging taking advantage of a pulsed source and hence time-of-flight methods is a new field unlocking yet unexplored methods and fields of applications. Within the last decade various novel imaging modalities have been developed taking advantage of wavelength resolution and in particular of scattering information amenable in addition to conventional real space image resolution. Hence, neutron imaging is not only revealing macroscopic inner structures or densities of different material compositions anymore, but also inhomogeneities of crystalline structure and phase, corresponding structural transitions as well as microscopic structural behavior in space and time and magnetic phenomena can be studied. An optimized imaging instrument at the ESS long pulse source will have the unique potential to cover
all these areas with highest efficiency and performance by taking advantage not only of the high time averaged flux but also the flexibility provided by the source time structure for a versatile instrument. However, taking advantage of such potential does not only require the instrument, as approved for construction and known as ODIN, but the corresponding development of method specific software tools, correction algorithms and quantification. Conventional and novel methods and their respective requirements shall be introduced.

15:45 End-to-end data-management facility at the faculty of Science - Brian Vinter, UCPH
The faculty of Science at UCHP has, in the past few years, build an infrastructure that allows researchers that use large scale imaging facilities, to manage their data end-to-end. Scientists may transfer data directly from a facility into UCPH, and may process them immediately, manually or through a workflow, and finally make archives and open-databases. The talk will introduce the motivation and the solutions that the solution offers and demonstrate how automated workflows may increase researcher productivity immensly.

16:15 CINEMAXe – workflow for 3D Modelling and Imaging of Material Microstructure - Jens W Andressen, DTU
DTU and KU will launch an e-course in 3D Modelling and Imaging of Material Microstructure in 2018. The course is based on our successful series of five-day summer schools, held in August 2015, 2016 and 2017.
We are now committed to transforming the lectures and material to an online course on Coursera. Our aim is to create an engaging course for both specialized PhD students and people from industry wanting to learn about high-level concepts and applications.
Besides video lectures and tutorials, the students learn techniques through hands-on exercises involving python programming implemented in Jupyter Notebooks. The notebooks are hosted and executed on JupyterHub servers, which were tested with 30 students on the most recent edition of the summer school. I will present the course program and show examples of the interactive sessions that the students are subjected to.

17:00 End of workshop