arXiv subjects

Henrik Birkedal

Publications and source records attributed to Henrik Birkedal.

4 recordsLinked to original sources

arXiv2026

The crystalline properties of silica biomorphs vary within and between morphologies

Silica-witherite biomorphs are a class of emergent materials, i.e. composite microstructures made of nanometric barium carbonate surrounded by amorphous silica. They form via co-precipitation of barium carbonate and siliceous species, and self-organize into a multitude of shapes with a distinct long-range order of the carbonate nanocrystals. However, the internal structural organization within and across different morphologies remains insufficiently resolved. Here, we use X-ray texture and diffraction tomography to create three-dimensional, spatially resolved maps of crystallographic orientation and structural parameters in silica-witherite biomorphs. At the sub-micron voxel level, all morphologies exhibit a crystallographic order consistent with a fiber texture around the c-axis. At larger length scales, however, the orientation field as well as crystallite size, crystallite shape anisotropy and the unit cell volume show systematic spatial variations. Leaf-like and helical morphologies contain defined directions along which structural parameters change systematically. Furthermore, we find recurring structural regimes with strong similarities between these morphologies. Conversely, coral-like morphologies are overall less textured and outside of the nucleation region we do not find clear structural regimes in the crystalline properties. These results provide a three-dimensional description of the internal organization of crystallites in silica-witherite biomorphs and establish a basis for systematically relating crystallographic organization to morphology.

cond-mat.mtrl-sci↗

arXiv2026

Direct detection of hydrogen reveals a new macroscopic crustal water reservoir on early Mars

The next great leap in Martian exploration will be the return of samples to Earth. To ensure the maximum scientific return from studying these samples, the development and utilisation of nondestructive analytical techniques are essential to enable early three-dimensional characterisation of their interiors. Neutron computed tomography is a powerful method in this context: it is highly sensitive to hydrogen and complements the more conventional X-ray computed tomography. Because the distribution and nature of hydrous phases are central to understanding the habitability, the climatic and geological evolution, and potential biosignatures of Mars, identifying hydrogenbearing phases in Martian crustal rocks is of particular importance. Using the only Martian crustal material available on Earth, the NWA 7034 meteorite and its pairs, we show that combined neutron and X-ray computed tomography enables non-destructive sample-wide mapping of hydrogen and reveals the distribution and petrographic contexts of hydrous phases. We identify hydrogen-rich iron oxyhydroxides within ancient igneous clasts, forming a macroscopic mineralogical water reservoir within the meteorite. These alteration assemblages closely resemble those observed in samples collected by the Perseverance rover in Jezero crater, where hydrated iron oxyhydroxides are also present. This similarity suggests that such phases may represent a widespread near-surface water reservoir on early Mars.

astro-ph.EP↗

arXiv2025

Disconnect to Connect: A Data Augmentation Method for Improving Topology Accuracy in Image Segmentation

Accurate segmentation of thin, tubular structures (e.g., blood vessels) is challenging for deep neural networks. These networks classify individual pixels, and even minor misclassifications can break the thin connections within these structures. Existing methods for improving topology accuracy, such as topology loss functions, rely on very precise, topologically-accurate training labels, which are difficult to obtain. This is because annotating images, especially 3D images, is extremely laborious and time-consuming. Low image resolution and contrast further complicates the annotation by causing tubular structures to appear disconnected. We present CoLeTra, a data augmentation strategy that integrates to the models the prior knowledge that structures that appear broken are actually connected. This is achieved by creating images with the appearance of disconnected structures while maintaining the original labels. Our extensive experiments, involving different architectures, loss functions, and datasets, demonstrate that CoLeTra leads to segmentations topologically more accurate while often improving the Dice coefficient and Hausdorff distance. CoLeTra's hyper-parameters are intuitive to tune, and our sensitivity analysis shows that CoLeTra is robust to changes in these hyper-parameters. We also release a dataset specifically suited for image segmentation methods with a focus on topology accuracy. CoLetra's code can be found at https://github.com/jmlipman/CoLeTra.

cs.CV↗

arXiv2022

Polychromatic neutron phase contrast imaging of weakly absorbing samples enabled by phase retrieval

We demonstrate the use of a phase retrieval technique for propagation-based phase contrast neutron imaging with a polychromatic beam. This enables imaging samples with low absorption contrast and/or improving the signal-to-noise ratio to facilitate e.g. time resolved measurements. A metal sample, designed to be close to a pure phase object, and a bone sample with canals partially filled with D2O were used for demonstrating the technique. These samples were imaged with a polychromatic neutron beam followed by phase retrieval. For both samples the signal-to-noise ratio were significantly improved and in case of the bone sample, the phase retrieval allowed for separation of bone and D2O, which is important for example for in situ flow experiments. The use of deuteration-contrast avoids the use of chemical contrast enhancement and makes neutron imaging an interesting complementary method to X-ray imaging of bone.

physics.med-ph↗