A research team from CEITEC BUT led by Markéta Kaiser participated in an international study to investigate the effect of protein intake on the fetus during pregnancy. The study was conducted on mice, or mouse embryos to be more precise. Thanks to the imaging possibilities offered by computed tomography in our CT Lab CEITEC, it was possible to create illustrative 3D models of cartilaginous tissue that clearly showed the differences between mouse embryos with different protein intakes.
At CT Lab CEITEC, researchers led by Prof. Jozef Kaiser focus on basic and applied research in collaboration with industry. The industrial partners are critical in applied research. Computed tomography is a non-invasive, non-destructive method that allows the detection of cracks or other defects on a wide variety of components so that if a company discovers that there is a problem somewhere, the CT Lab staff can detect and describe the problem without the company having to cut up or otherwise dismember the expensive product or component. The knowledge gained in this way in turn advances the researchers’ understanding of the phenomenon.
When it comes to basic research, the researchers are involved in so-called reverse engineering, where professional data display and analysis programs, which are primarily designed for mechanical engineering, are used in biology. Using advanced methods, they are able to image not only bones but also soft tissues such as cartilage or muscles.
This is also the subject of the recently published study <a href="https://www.nature.com/articles/s41467-024-46030-3">“The level of protein in the maternal murine diet modulates the facial appearance of the offspring via mTORC1 signaling”</a>, which appeared in prestigious Nature Communications.
Scientists from the Laboratory of X-ray Computational Micro and Nanotomography at CEITEC, led by Markéta Kaiser, have been collaborating on this study since 2018. The multidisciplinary international team was coordinated by prof. Andrei Chagin from the University of Gothenburg, Sweden. “Since then, dozens of samples have been analysed in our laboratory. Our task was not only to create 3D models of cartilage tissue, but also to analyze these 3D models. Here, we have the advantage of being a technical university and working closely with industry in the lab, and so we are using just industrial tools in biological research in a somewhat unconventional way, which is not quite so common and can bring interesting insights into these interdisciplinary studies,” says Markéta Kaiser.
We now know that facial shape and features in mammals are influenced not only by genetics but also by the mother’s lifestyle during pregnancy. It is known that, for example, the consumption of alcohol or other addictive substances can lead to facial abnormalities or developmental disorders. The present study reveals a possible new link between the offspring’s face and pregnancy lifestyle, specifically protein intake during pregnancy.
The study was conducted on mice or mouse embryos. “Specifically, the researchers focused on the mTORC1 signaling pathway, and whether protein content in pregnancy diet affects mTORC1 and plays a role in facial bone structure. Pregnant mice were fed a low- and high-protein diets, then the effect on the aforementioned signalling pathway was investigated and facial development in mouse embryos was compared,“ Markéta Kaiser explains the methodology.
It was at this point that X-ray computed tomography analysis played a key role. Scientists from CEITEC BUT created 3D models of the cartilaginous tissue that is the basis for the subsequent development of the face. The differences between embryos from mice having different protein diets were noticeable, albeit very subtle. For example, an enlarged nasal cavity or an elongated jaw were observed. Therefore, the study shows that the amount of protein in the diet can affect mTORC1 and thus the subsequent musculoskeletal development of the individual.
This collaboration certainly does not end with the published article. The researchers already have more projects planned that could help unravel other processes related to more than just facial development.
<img src="https://storage.googleapis.com/labrulez-bucket-strapi-h3hsga3/CEITEC_Prijem_bilkovin_matky_behem_tehotenstvi_muze_ovlivnit_vyvoj_obliceje_novorozence_e682d5cd06.jpg" alt="CEITEC BUT: 3D cartilage tissue models of mouse embryos created using X-ray computed microtomography. The different colours of the models represent different protein diets."/>

Mother’s protein intake during pregnancy can affect newborn’s facial development

In November 2025, the X-ray Computed Tomography Laboratory at CEITEC Brno University of Technology will be expanded with a new state-of-the-art phoenix v|tome|x L450 computed tomography scanner from Waygate Technologies. This high-performance device will expand the laboratory's already unique and extensive range of equipment and take research and industry collaboration to a whole new level.<img src="https://storage.googleapis.com/labrulez-bucket-strapi-h3hsga3/CTLAB_CEITEC_Novy_vypocetni_tomograf_odhali_dutiny_a_vady_u_velkych_prumyslovych_dilu_f2c8ce5cdb.jpg" alt="CTLAB CEITEC: New computed tomography scanner detects cavities and defects in large industrial parts" srcset="https://storage.googleapis.com/labrulez-bucket-strapi-h3hsga3/thumbnail_CTLAB_CEITEC_Novy_vypocetni_tomograf_odhali_dutiny_a_vady_u_velkych_prumyslovych_dilu_f2c8ce5cdb.jpg 235w," sizes="100vw" width="1600" height="1067">The Phoenix v|tome|x L450 is designed for non-destructive testing and measurement of larger and heavier samples, which until now have been almost impossible to inspect using this technology. Computed tomography (CT) is an imaging method that uses X-rays to create detailed 3D images. Using a computer, a large number of 2D images (up to 5,000) are assembled using tomographic reconstruction to create a 3D model that shows the internal structure of the sample without damaging it. The method is best known in healthcare, where it allows doctors to see the body in cross-sections, and diagnose various medical conditions.At the X-ray Computed Tomography Laboratory at CEITEC BUT (<a target="_blank" rel="noopener noreferrer" href="https://ctlab.ceitec.cz/about-us/?_gl=1*13y0662*_ga*OTU0MTYzNzc1LjE3Mzg5MTQ2MDk.*_ga_L74QG6CYBE*czE3NjQwNTc2Nzkkbzg3JGcxJHQxNzY0MDU4MTg2JGo2MCRsMCRoMA..">CTLAB CEITEC</a>), we specialize in contract research, among other things. We provide expertise to industrial partners from a wide range of sectors, from automotive and aviation to defense and aerospace. “We can detect hidden defects, whether cracks or porosity, check internal geometric structures, inspect electrical assemblies, and more. This detailed information helps optimize production processes, ensure reliability, improve performance, and extend product life,” says Jozef Kaiser, head of <a target="_blank" rel="noopener noreferrer" href="https://www.ceitec.eu/advanced-instrumentation-and-methods-for-materials-characterization/rg6">Advanced Instrumentation and Methods for Materials Characterization</a> research group to which the laboratory belongs.The new computed tomography scanner will enable applications for the defense and aerospace industries, as well as research into new materials and structured composites.“With the L450 device, we gain access to breakthrough technology in the field of non-destructive analysis of materials and components – especially for large, dense, or composite objects, where existing technologies have reached their limits,” adds Tomáš Zikmund, head of the laboratory.<h4>What the phoenix v|tome|x L450 can do</h4>The phoenix v|tome|x L450 CT system stands out thanks to its high-energy X-ray source of up to 450 kV and robust mechanical design, making it ideal for:<ul><li>analyzing large and heavy parts, including metal castings, composites, or plastic assemblies with dense components (Maximum scan volume: Ø800 x 1250 mm, Maximum sample height: 2 m, Maximum sample weight: 200 kg, Minimum focal spot: 63 µm),</li><li>defectoscopy, i.e. detection of internal defects, cracks, cavities, and delaminations without destroying the sample,</li><li>3D metrology, i. e. accurate, non-contact measurement of internal geometry with high resolution,</li><li>digital archiving, and reverse engineering.</li></ul><img src="https://storage.googleapis.com/labrulez-bucket-strapi-h3hsga3/CTLAB_CEITEC_Novy_pocitacovy_tomograf_Phoenix_vtomex_L450_2dfef07a4d.jpg" alt="CTLAB CEITEC: Phoenix v|tome|x L450" srcset="https://storage.googleapis.com/labrulez-bucket-strapi-h3hsga3/thumbnail_CTLAB_CEITEC_Novy_pocitacovy_tomograf_Phoenix_vtomex_L450_2dfef07a4d.jpg 235w," sizes="100vw" width="1600" height="1067"><h4>Tested on a student formula car</h4>Before purchase, the L450 device was tested directly at the manufacturer’s premises in Germany – on a BUT student racing car project (TU Brno Racing team). The device was used to perform a detailed 3D CT scan of the carbon monocoque of the Dragon e15 racing car. The results of the analysis confirmed not only the ability of the device to detect subtle defects in composite structures, but also its potential in the development of highly safe and lightweight designs.<h4>Significant expansion of competencies</h4>The new device will complement the existing equipment of the computed tomography laboratory at CEITEC BUT, which is already unique in Central Europe. “The L450 will enable us not only to expand our research activities, but also to offer advanced services to companies and partners working with large or complex components,” concludes Jozef Kaiser.

New computed tomography scanner detects cavities and defects in large industrial parts

LA-ICP-MS imaging, particularly in TOF mode, provides rapid, high-resolution elemental maps of complex geological materials, but quantitative accuracy is often limited by heterogeneous mineralogy. This study introduces a workflow using UMAP for unsupervised dimensionality reduction combined with k-means clustering to segment mineral phases directly from per-pixel elemental data.Cluster compositions are matched to known minerals based on stoichiometry, allowing phase-specific normalization and reducing quantification errors by up to 60%. Validated on dawsonite-bearing sandstones, the approach enables consistent phase-resolved geochemical comparisons across depth profiles and establishes UMAP as a practical tool for accurate, interpretable multielement imaging.<blockquote><h4>The original article</h4><h4><a target="_blank" rel="noopener noreferrer" href="https://pubs.acs.org/doi/10.1021/acs.analchem.5c05398">Mineral Phase-Resolved Quantification in LA-ICP-MS Imaging</a></h4>Barbara Umfahrer, Jakub Buday, Pavel Pořízka, Jozef Kaiser, Paolo S. Garofalo*, and Detlef Günther*Anal. Chem. 2026, 98, 1, 581–589<a target="_blank" rel="noopener noreferrer" href="https://doi.org/10.1021/acs.analchem.5c05398">https://doi.org/10.1021/acs.analchem.5c05398</a><h6>licensed under <a target="_blank" rel="noopener noreferrer" href="https://creativecommons.org/licenses/by/4.0/">CC-BY 4.0</a></h6></blockquote>Selected sections from the article follow. Formats and hyperlinks were adapted from the original.Since its introduction in 1985, <a href="javascript:void(0);">(1)</a> Laser Ablation-Inductively Coupled Plasma-Mass Spectrometry (LA-ICP-MS) has matured into a routine technique for both qualitative and quantitative elemental analysis, finding use in diverse fields such as geology, <a href="javascript:void(0);">(2−4)</a> biology, <a href="javascript:void(0);">(5,6)</a> materials science, <a href="javascript:void(0);">(7)</a> and cultural heritage. <a href="javascript:void(0);">(8,9)</a> Especially in comparison to liquid analysis techniques, the method excels through minimal sample preparation, negligible sample destruction, and low limits of detection, which enable trace element analysis. However, besides the availability of standard reference materials, a longstanding challenge of quantitative LA-ICP-MS is the need for a suitable internal standard to account for variations in ablation rate and transport efficiency, as well as instrumental drift. <a href="javascript:void(0);">(10)</a> Developed for applications in geological matrices, this requirement was elegantly circumvented by Liu et al. <a href="javascript:void(0);">(11)</a> through 100 wt % normalization, which assumes that all detected elements form oxides summing up to 100 wt %. This approach was conceptually inspired by the summed-spectrum normalization calibration procedure introduced by Leach and Hieftje in 2000, which emphasized the use of total signal intensities for robust internal standardization. <a href="javascript:void(0);">(12)</a> The 100 wt % normalization strategy was subsequently extended to the bulk analysis of carbonates. <a href="javascript:void(0);">(13)</a>Concurrently, the advent of time-of-flight (TOF) mass analyzers and the development of fast-washout laser ablation cells <a href="javascript:void(0);">(14−16)</a> have transformed LA-ICP-MS into a powerful imaging tool, capable of generating spatial maps across a variety of sample matrices at high data acquisition rates. <a href="javascript:void(0);">(17−19)</a> Furthermore, in comparison to, e.g., X-ray fluorescence spectroscopy (XRF) or Scanning Electron Microscopy with Energy-Dispersive X-ray Spectroscopy (SEM-EDS), LA-ICP-MS offers a high depth resolution of approximately 0.2 μm on glass, enabling precise measurement of fine-scale structures not only laterally but also in depth. <a href="javascript:void(0);">(18)</a> However, as the volume and complexity of these elemental images have grown according to the progressive advancement of the method itself, now consisting of gigabytes of data produced within minutes, manual data interpretation has become increasingly impractical. To uncover meaningful details and patterns, researchers have applied multivariate methods such as Principal Component Analysis (PCA) and k-means clustering to reduce dimensionality and delineate distinct regions within the images. <a href="javascript:void(0);">(20−23)</a> The nonlinear dimensionality reduction technique Uniform Manifold Approximation and Projection (UMAP), <a href="javascript:void(0);">(24)</a> which supports data interpretation with its capability to capture and separate subtle local structures, has first been employed to LA-ICP-MS bulk analysis data in 2024, where it was used to classify pyrite samples and discriminate between Pb and Zn deposit types. <a href="javascript:void(0);">(25,26)</a> In combination with elemental imaging, UMAP has recently been applied to biological tissue data sets as an exploratory analysis tool, <a href="javascript:void(0);">(27)</a> further enhancing the ability to visualize and interpret complex elemental distributions.Despite these advances, no existing approach combines the unsupervised segmentation of mineral phases by UMAP with phase-specific normalization. In this study, we introduce a UMAP-driven workflow that (1) automatically segments distinct mineralogical phases within fine-grained geological matrices and (2) applies phase-specific 100 wt % normalization (e.g., oxide <a href="javascript:void(0);">(11)</a> or carbonate <a href="javascript:void(0);">(13)</a>), thereby improving quantification accuracy and enabling phase-specific quantitative interpretation of mineral distributions in complex matrices.<h4>Materials and Methods</h4><h5>LA-ICP-TOFMS Imaging</h5>LA-ICP-TOFMS analyses were performed using a 193 nm ArF excimer laser (GeoLas C, Lambda Physik, Göttingen, Germany) coupled to an ICP-TOFMS instrument (icpTOF 2R, TOFWERK AG, Thun, Switzerland). Ablation was conducted in the modified parallel flow ablation cell (MPFAC) <a href="javascript:void(0);">(16,34)</a> under a He atmosphere (99.999%, Linde Gas Schweiz AG, Dagmersellen, Switzerland) with Ar (99.996%, Linde Gas Schweiz AG) used as the carrier gas. A piezoelectrically driven xyz translational stage (1 nm resolution, SmarAct GmbH, Oldenburg, Germany) enabled precise sample positioning. Laser, stage movement, and ICP-TOFMS data acquisition were synchronized using a custom-built imaging control system. <a href="javascript:void(0);">(19)</a> The system was optimized each day based on maximum sensitivity while ensuring a 238U+/232Th+ intensity ratio of 1.0 to 1.1 on NIST SRM 610 and a 232Th16O+/232Th+ rate of below 0.5%.<h5>LIBS Imaging</h5>The LIBS analyses were performed on a Firefly system (Lightigo, Czech Republic). The laser (1064 nm, 5 mJ, pulse duration ∼10 ns, diameter 6 mm, operating at 50 Hz) was focused on the sample with a singlet lens (focal length 50 mm) to a 30 μm spot. The spatial resolution of the acquired elemental maps, with a size of about 16 mm × 26 mm, was 20 μm. Plasma radiation is collected by a multifiber configuration with a collimator and an optical fiber (400 μm core diameter), which delivered the light to several Czerny–Turner spectrometers (resolution 0.25–0.4 nm). The spectral range of the spectrometers is 180–850 nm. The radiation is collected 1.5 μs after ablation, with a gate width of 15 μs.Data processing was done with Python 3.11.8 in the Spyder environment and included background correction and per-pixel z-score normalization.<h4>Results and Discussion</h4><h5>LIBS and XRD Measurements</h5>To assess whether carbonate-based normalization <a href="javascript:void(0);">(13)</a> is warranted, first, large-scale LIBS elemental intensity maps of the samples across about a third of the thin sections were acquired. <a target="_blank" rel="noopener noreferrer" href="https://pubs.acs.org/doi/10.1021/acs.analchem.5c05398#fig1">Figure 1</a> reveals the results for “Layer 2” of the Poggio del Gatto sandstone for selected elements. A complete mapping of all four thin sections can be found in <a target="_blank" rel="noopener noreferrer" href="https://pubs.acs.org/doi/suppl/10.1021/acs.analchem.5c05398/suppl_file/ac5c05398_si_001.pdf">Figures S3 and S6</a> in the Supporting Information. The maps reveal that C is strongly enriched in fracture veins as well as in finer-grained domains within the matrix and correlates with Ca, indicating carbonate minerals rather than oxides. Bulk XRD analysis identifies calcite alongside quartz, plagioclase, white mica, and kaolinite. These complementary observations confirm calcite as a major phase and thus justify the need for performing the 100 wt % normalization on a carbonate basis for pixels representing calcite.<img src="https://storage.googleapis.com/labrulez-bucket-strapi-h3hsga3/Anal_Chem_2026_98_1_581_589_Figure_1_LIBS_elemental_intensity_maps_of_the_data_set_from_Layer_2_of_the_Poggio_del_Gatto_sandstone_mounted_on_a_microscope_slide_73ada1b7b8.jpg" alt="Anal. Chem. 2026, 98, 1, 581–589: Figure 1. LIBS elemental intensity maps of the data set from “Layer 2” of the Poggio del Gatto sandstone mounted on a microscope slide, depicting calcite veins and fine-grained structures within the matrix, underlining the need for carbonate-based normalization for specific pixels. Emission lines for evaluation are shown in the corresponding subfigure title. The epoxy used to fix the thin section on the slide is visible along the margin." srcset="https://storage.googleapis.com/labrulez-bucket-strapi-h3hsga3/thumbnail_Anal_Chem_2026_98_1_581_589_Figure_1_LIBS_elemental_intensity_maps_of_the_data_set_from_Layer_2_of_the_Poggio_del_Gatto_sandstone_mounted_on_a_microscope_slide_73ada1b7b8.jpg 245w," sizes="100vw" width="2088" height="414"><h5>Renormalization Based on Carbonates</h5>Applying an oxide-based normalization to carbonate samples will introduce composition-dependent biases that cannot immediately be generalized as a singular numerical value. The difference between the molecular masses of carbonate CO32– and its oxide equivalent O2– yields an intrinsic bias on the order of a factor of about two, depending on the stoichiometry. Once this data is renormalized to a total of 100 wt %, the magnitude of the error becomes highly dependent on the specific elemental composition of the sample. An exemplary calculation illustrating this effect for a representative calcite pixel is provided in <a target="_blank" rel="noopener noreferrer" href="https://pubs.acs.org/doi/suppl/10.1021/acs.analchem.5c05398/suppl_file/ac5c05398_si_001.pdf">Table S2</a>, yielding a relative error of 79%. In practice, this error varies across the image according to the local phase composition. <a target="_blank" rel="noopener noreferrer" href="https://pubs.acs.org/doi/10.1021/acs.analchem.5c05398#fig6">Figure 6</a> illustrates the impact of normalization approaches for layer 2 of the samples as well as the spatial distribution of the resulting relative error. The figure presents Ca and Si elemental maps in two rows, while the first three columns compare different normalization strategies: normalization based on all oxides, normalization based on all carbonates, and a pixel-wise normalization guided by mineral phase identification, as outlined in the workflow proposed in this study. The final column displays the relative error between the bulk oxide- or carbonate-based normalization and the pixel-wise, phase-informed normalization.<img src="https://storage.googleapis.com/labrulez-bucket-strapi-h3hsga3/Anal_Chem_2026_98_1_581_589_Figure_6_Comparison_of_normalization_strategies_for_Ca_and_Si_LA_ICP_TOFMS_elemental_maps_in_Layer_2_ce3a0c1a5c.jpg" alt="Anal. Chem. 2026, 98, 1, 581–589: Figure 6. Comparison of normalization strategies for Ca and Si LA-ICP-TOFMS elemental maps in “Layer 2” of the Poggio del Gatto sandstone. Rows show Ca (top) and Si (bottom); columns compare (1) oxide-based normalization, (2) carbonate-based normalization, (3) pixel-wise, phase-informed normalization (this study), and (4) relative error between the first two and the phase-informed approach." srcset="https://storage.googleapis.com/labrulez-bucket-strapi-h3hsga3/thumbnail_Anal_Chem_2026_98_1_581_589_Figure_6_Comparison_of_normalization_strategies_for_Ca_and_Si_LA_ICP_TOFMS_elemental_maps_in_Layer_2_ce3a0c1a5c.jpg 202w," sizes="100vw" width="2088" height="1622">For Ca (in this matrix primarily associated with calcite, a dominant carbonate phase in our samples), the relative error depicted quantifies the effect of applying oxide-based normalization across the entire image. This results in a systematic overestimation of the Ca concentration by approximately 60%. Conversely, for Si─in this matrix found mainly in quartz and various plagioclase phases─the error is shown for applying carbonate-based normalization across an oxidic matrix. In such cases, the Si concentrations are underestimated by roughly 10% to 40%, depending on the mineral phase.These findings highlight that using a single normalization scheme across heterogeneous samples introduces substantial quantification errors. To mitigate this, the pixel-wise phase segmentation enabled by the proposed UMAP-based workflow offers a significant advantage, especially in fine-grained matrices. In short, phase-aware, pixel-wise normalization combines the strengths of both normalization approaches and proves to be highly beneficial.<h4>Conclusion</h4>This study demonstrates that phase-specific normalization, guided by unsupervised phase segmentation of LA-ICP-MS images, can substantially improve the accuracy of elemental quantification in fine-grained geological matrices. Rather than relying on a single global assumption (e.g., all oxides or all carbonates), the workflow adapts normalization to the actual mineralogy, minimizing systematic errors and enabling phase-resolved geochemical interpretation.The segmentation step, based on UMAP embedding and k-means clustering, is fully unsupervised. However, the subsequent matching of clusters to specific minerals still relies on a limited set of preidentified stoichiometries and is therefore not yet fully unsupervised. Expansion to larger stoichiometric databases could help overcome this limitation and move toward automated mineral classification. A key challenge remains the similarity of cation ratios in certain mineral groups, which may still require complementary data to ensure robust assignments.Overall, beyond the specific case of the Poggio del Gatto sandstone presented in this study, this work provides a framework for quantitative elemental imaging of complex geological samples and is readily transferable to other contexts that require spatially resolved composition-aware normalization. This contribution adds a practical implementation to a growing literature that already uses multivariate statistics to unlock the immense information depth of LA-ICP-TOFMS.

Mineral Phase-Resolved Quantification in LA-ICP-MS Imaging

Integrated laser ablation and computed tomography: detecting the chemical footprint of microplastics and reconstructing 3D biological tissues

A Method for the Detection of Tire Wear Microplastics in Zebrafish Guts by Laterally Resolved LA-ICP-MS-Based Elemental Fingerprinting and Chemometrics

Article

Product

Events

Webinars

Application

Academy

Video

Poster

Scientists

Plastics

Environment

Food

Water

Health

Presentation

Scientific article

Interview

Toxicology

Popularization

Science and research

Software

Covid-19

Career

Laboratories

Various

Zaměření

Type of News

Focus

Je výzkumné centrum zaměřené na oblast živých věd, pokročilých materiálů a technologií.  Vznikl v roce 2011 na základech šesti významných brněnských univerzit a výzkumných institucí. Od doby jeho nedávného vzniku se daří budovat významné vědecké centrum, a již nyní se řadí mezi špičkové evropské instituce.

CEITEC

Vše ze světa analytické chemie na jednom místě. Propojujeme lidi při řešení jejich problémů. U nás naleznete všechny potřebné informace snadno, rychle a přehledně. Přestaňte hledat a začněte nacházet.

LabRulez

18th Mass Spectrometry School in Biotechnology and Medicine

28th International Symposium on Advances in Extraction Technologies (ExTech 2026)

The 22nd Biennial National Atomic Spectroscopy Symposium (BNASS 2026)

End-to-end PAT: Connecting advanced analytics and data across the bioprocess lifecycle

Simplifying Through-Container Testing with Raman Spectroscopy in the Pharmaceutical Industry

Bruker Fluorescence Microscopy Solutions: Automated Imaging and Light-Sheet Microscopy