Simon Zabler’s research while affiliated with Deggendorf Institute of Technology and other places

Optimizing phase‐contrast micro‐computed tomography (µCT) for a given object is not trivial if the radiation is polychromatic and the object multi‐material. This study demonstrates how an optimal combination of propagation distance and mean energy (set by attenuation filters) may be derived for such an object (an electromotor scanned on beamline BM18 at ESRF in Grenoble, France). In addition to appropriate image quality metrics, it is mandatory to define a task. In that respect, raising Emean from 100 keV to 164 keV mitigates beam hardening by metal parts, yet raising Emean further to 230 keV deteriorates CNR² (where CNR is contrast‐to‐noise ratio) due to higher image noise. Propagation distances between d = 2 m and 25.3 m are evaluated crosswise with energy. While longer propagation distances generally yield higher CNR², shorter distances appear favorable when discerning plastic near metal parts. SNR² (where SNR is signal‐to‐noise ratio) power spectra and modulation transfer (MTF) are evaluated independently from two‐dimensional projections supporting volume image analysis for which image sharpness depends strongly on the digital filters (Paganin and Wiener) which are applied along with filtered back‐projection. In summary, optimizing synchrotron µCT scans remains a very complex task which differs from object to object. A physically accurate model of the complete imaging process may not only allow for optimization by simulation but also ideally improve CT image reconstruction in the near future.

Conventional industrial computed tomography (CT) systems are constrained in their choice of acquisition trajectories due to their mechanical design. These systems are very precise instruments since they do only move on primarily highly accurate rotational stages. In order to be able to scan an arbitrary Region of Interest (ROI), regardless of the position, size and weight of the specimen, conventional industrial robots can be used as flexible 6 degrees of freedom (DOF) manipulators. For example in a twin robot computed tomography system, acquisition geometries with arbitrary tool poses can be realized. In scientific applications, the quality of the CT volume image is of primary interest, whereas in an industrial environment it is often a matter of balancing quality and acquisition time. Common industrial robots cannot achieve the required positional accuracy without calibration to generate an ideal reconstruction. In the presented study, methods for the geometric correction of CT scans are compared. Image based correction is compared to general machine calibration and full pose tracking by laser trackers. Image quality metrics such as the Modulation Transfer Function, Shannon entropy and Tenengrad variance are utilized to evaluate and compare the reconstruction quality of the various correction and calibration approaches. The assessment of the reconstruction quality revealed a comparable reconstruction quality between the approaches, with the machine calibration approach emerging as one of the best, while also reducing the time-intensive correction overhead.

The size of 4D tomography datasets acquired at synchrotron or neutron imaging facilities can reach several terabytes, which presents a significant challenge for their evaluation. This paper presents a framework that allows a compressed dataset to be kept in memory and makes it possible to evaluate and manipulate the dataset without requiring enough memory to decompress the entire dataset. The framework enables the compensation of imaging artifacts, including the compression artifacts of the 4D dataset, through the integration of neural networks. The reduction of imaging artifacts can be performed at the imaging facility or at the user's home institution. This framework reduces the computational burden on the computing infrastructure of large synchrotron and neutron facilities by allowing end users to process datasets on their institution's computers. This is made possible by compressing TBs of data to less than 128 GB, allowing powerful PCs to process TBs of 4D tomography data.

Low-dose CT scans are very fast, but they feature strong pixel noise which, in turn, invites for efficient denoising the same way that strong image blur invites for sharpening the images. Wiener-deconvolution serves as a starting point, combining these two operations in a linear Fourier filter. It may apply in 2D or 3D and requires merely estimates of the Signal-to-Noise Ratio SNR2 as well as the system Modulation Transfer Function (MTF). Meanwhile, any linear filter is unable to adapt to local contrast variations in the images. Therefore, convolutional neural networks (CNN) such as MSDnet or UNet can significantly improve reconstructions from low-dose CT scans thanks to their non-linear nature. The Noise2Inverse framework allows for training these CNN using a split-and-merge of two subsets generated from one raw CT dataset. This study investigates denoising by N2I with respect to the input data quality, the CNN network architecture and the fidelity of the features in the resulting denoised volume images. Unlike Wiener filtering, N2I is far more effective in removing pixel noise. Yet, these models give the vague impression that they might generate ghosts of small features (pores or spots) in the object. Image quality metrics like Fourier Shell Correlation and SNR2 power spectra show clear improvement. In particular, Noise2Inverse preserves the signal power spectrum of all scans reliably. Nevertheless, locally inspecting image features must remain a core technique when the CNN models are applied, i.e. for industrial inspection.

Transparent crystalline scintillators such as cerium-doped YAG or LuAG are widely used in X-ray imaging for the indirect detection of X-rays. The application of reflective coatings on the front side to improve the optical gain is common practice for flat panel detectors with CsI or Gd2O2S powder scintillators but still largely unknown for crystalline scintillators such as LuAG. This work shows experimentally and quantitatively how a black and reflective coating on the X-ray side of a 2 mm LuAG:Ce scintillator improves the image quality compared to a 2 mm LuAG:Ce scintillator without a coating. The measurements have been done for two different distances, with 2 m and 29.7 m on the BM18 beamline of the European Synchrotron. The Modulation Transfer Function (MTF) and the Signal-to-Noise-Ratio (SNR²) power spectrum as well as contrast-to-noise ratio are used for comparing image quality. Propagation-based phase contrast strongly enhances the SNR² amplitudes (gain ≈10 from 2 m to 29.7 m object-detector distance) of the raw images’ spectrum independent of the scintillator coating. For both detector positions, the reflective coating is able to raise SNR² by up to 80% through the improved optical gain, while black coating does the opposite (decrease SNR² by 20%) with respect to no coating. With the tested optical setups, changes in MTF /sharpness between the coatings are minor. Comparing CNR² in CT scans of a multi-material sample, in this case an electric motor, we observe the reflective coating yielding better material contrast for plastic and air. Application and effect of Wiener-deconvolution, along with Paganin-type phase retrieval, are also discussed in the context of CT image quality.

Background The stability of implant-abutment connection is crucial to minimize mechanical and biological complications. Therefore, an assessment of the microgap behavior and abutment displacement in different implant-abutment designs was performed. Methods Four implant systems were tested, three with a conical implant-abutment connection based on friction fit and a cone angle < 12 ° (Medentika, Medentis, NobelActive) and a system with an angulated connection (< 40°) (Semados). In different static loading conditions (30 N − 90º, 100 N − 90º, 200 N − 30º) the microgap and abutment displacement was evaluated using synchrotron-based microtomography and phase-contrast radioscopy with numerical forward simulation of the optical Fresnel propagation yielding an accuracy down to 0.1 μm. Results Microgaps were present in all implant systems prior to loading (0.15–9 μm). Values increased with mounting force and angle up to 40.5 μm at an off axis loading of 100 N in a 90° angle. Conclusions In contrast to the implant-abutment connection with a large cone angle (45°), the conical connections based on a friction fit (small cone angles with < 12°) demonstrated an abutment displacement which resulted in a deformation of the outer implant wall. The design of the implant-abutment connection seems to be crucial for the force distribution on the implant wall which might influence peri-implant bone stability.

Background: Correct individual tonotopic frequency stimulation of the cochlea plays an important role in the further development of anatomy based cochlear implantation. In this context frequency specific fitting of the basal electrode contact with a normal insertion depth can be difficult since it is often placed in a frequency range higher than 10 kHz and current audio processors only stimulate for frequencies up to 8.5 kHz due to microphone characteristics. This results in a mismatch of the high frequencies. Therefore, this study represents a proof of concept for a tonotopic correct insertion and aims to develop an algorithm for a placement of the basal electrode below 8.5 kHz in an experimental setting. Methods: Pre- and postoperative flat-panel volume CT scans with secondary reconstructions were performed in 10 human temporal bone specimens. The desired frequency location for the most basal electrode contact was set at 8.25 kHz. The distance from the round window to the position where the basal electrode contact was intended to be located was calculated preoperatively using 3D-curved multiplanar reconstruction and a newly developed mathematical approach. A specially designed cochlear implant electrode array with customized markers imprinted on the silicone of the electrode array was inserted in all specimens based on the individually calculated insertion depths. All postoperative measurements were additionally validated using an otological planning software. Results: Positioning of the basal electrode contact was reached with only a small mean deviation of 37 ± 399 Hz and 0.06 ± 0.37 mm from the planned frequency of 8.25 kHz. The mean rotation angle up to the basal electrode contact was 51 ± 5 °. In addition, the inserted electrode array adequately covered the apical regions of the cochleae. Conclusion: Using this algorithm, it was possible to position the basal electrode array contact in an area of the cochlea that could be correctly stimulated by the existing speech processors in the context of tonotopic correct fitting.

For about a decade, the Thünen Institute of Wood Research has been analyzing wood species in charcoal and briquettes. The anatomical analyses can be used to check whether the species declarations of the product samples are correct. In addition, the origin information on the distribution areas of the specific taxa can be scrutinized (e.g., tropical vs. non-tropical). The investigations make an important contribution to quality assurance for the producers and traders of the products, and also serve to protect consumers. In the context of the studies, it has been observed that more raw and residual materials, such as non-timber forest products (NTFPs), are becoming established on the charcoal substitutes market. Anatomical studies are currently being carried out to characterize the structure of important NTFPs, such as coconut shells, olive or mango kernels, in order to describe them and reliably identify them using 3D-Refelcted-Light- Microscopy (3D-RLM) and nano-CT in practical applications for market analyses