Image from the NFDI4BIOIMAGE Calendar Cover 2025. The image is a visualization showing the integration of multimodal data including a spinning disk confocal image and gene expression data from a spatial transcriptomic experiment on a human medulloblastoma sample. The microscopy image of the tissue with the nuclei in white has been overlayed with the result of the cell segmentation colored according to the assigned cell type (immune cells: red, stromal cells: violet, brain cells: cyan/blue, tumor cells: green). A subset of transcripts for three genes whose expression varies across the different cell types in the tissue have been represented as colored dots (CD4 (immune cells): red, PTCH1 (tumor cells): green, AQP4 (brain cells): blue). Image Metadata (using REMBI template): Study Study description Comparison of spatial transcriptomics technologies for medulloblastoma cryosection Study type Spatial Transcriptomics (Xenium) on medulloblastoma cryosections Study Component Imaging method Xenium and Spinning disk confocal microscopy Study component description Datasets with raw and processed data from the study "Comparison of spatial transcriptomics technologies for medulloblastoma cryosections" including Xenium and spinning disk confocal microscopy data Biosample Identity MB266 Biological entity Human cerebellum from a patient with Medulloblastoma with extensive nodularity Organism Homo sapiens Specimen Experimental status Patient sample Preparation method 10 µm cryosections were acquired using the cryostar NX50 with a cutting temperature of -15 °C. Tissues were section in 10 µm slices and four samples were placed on one Xenium slide. Subsequently, the tissue was fixed with PFA according to the manufacture´s protocol. Tissues were permeabilized with SDS, incubated in 70% ice cold methanol and washed with PBS. Hybridization of the human generic brain panel with 70 add-on genes (Supplementary Dataset 1) was performed at 50°C in a Bio-Rad C1000 touch cycler for 20 hours. Washing, ligation and amplification steps were carried out according to the manufacturer’s instructions. ROIs were selected according to the tissue area excluding non-tissue covered tiles. Each transcript was imaged in a bright state five times across 60 cycle-channels (15 cycles x 4 channels). After the run on the Xenium analyzer slides were removed and buffer exchanged with PBS-T for further storage at 4°C. Signal/contrast mechanism Fluorescence Channel 1 – content DAPI Channel 1 – biological entity Nuclei (DNA) Image acquisition Instrument attributes Imaging of RNAscope samples and reimaging of Xenium slides by SDCM was conducted on an Andor Dragonfly 505 spinning disk confocal system equipped with a Nikon Ti2-E inverted microscope and a CFI P-Fluor 40X/1.30 oil objective or a Plan Apo 60x/1.40 oil objective. Multicolor images were acquired with the following laser lines 405 nm (DAPI), 488 nm (Alexa 488, eosin), 561 nm (Atto 550), 637 nm (Atto 647) 730nm (Alexa 750). Image acquisition parameters Images were recorded at 16-bit depth and with 1024x1024 pixels dimensions (pixel size: 0.217 µm) using an iXon Ultra 888 EM-CCD camera. The region of interest was selected based on the DAPI signal and 50 z-slices were acquired with a step size of 0.4 µm (20 µm z-range) per field of view (FOV). Tiles were imaged with a 10% overlap to ensure accurate stitching. Image data Type Figure Format & compression PNG Size description 8800x8788+0+0 pixels (Primary image) Pixel/voxel size description 0.217 µm (Primary image) Channel information RGB Image processing method Tiles were imaged with a 10% overlap to ensure accurate stitching. Subsequently, a flatfield-correction was conducted based on the DAPI channel and stitching and registration of the tiles was conducted with Fiji. First, SDCM image stacks were subjected to a maximum intensity projection, followed by flat field and chromatic aberration correction using a custom script. Next, image tiles were stitched using the “Grid/Collection Stitching” plugin. DAPI images from SDCM were registered to MC or Xenium widefield images using “Register Virtual Stack Slices” with Affine feature extraction model and the Elastic bUnwarpJ splines registration model. In case of further staining, images were transformed via Transform Virtual Stack slices employing the transformation file of the DAPI registration. Image Correlation Spatial and temporal alignment The region of interest was selected based on the DAPI signal and 50 z-slices were acquired with a step size of 0.4 µm (20 µm z-range) per field of view (FOV). Tiles were imaged with a 10% overlap to ensure accurate stitching. Subsequently, a flatfield-correction was conducted based on the DAPI channel and stitching and registration of the tiles was conducted with Fiji (https://github.com/RippeLab/MBEN/tree/main/stitching) (https://github.com/RippeLab/MBEN/tree/main/Registration). Related images and relationship MB266-morphology_mip.ome.tif at https://www.ebi.ac.uk/biostudies/bioimages/studies/S-BIAD1093 Analysed data Analysis result type Figure Data used for analysis MB266-transcripts.csv.gz, MB266-transcripts.csv.gz at https://www.ebi.ac.uk/biostudies/bioimages/studies/S-BIAD1093 Analysis method and details Most of the analysis and visualization (including tidyverse, data.table, ggridges R packages) was done in R 4.2.2. Raw data were processed using technology-specific corporate pipelines (custom pipeline was used for MC). For each technology Seurat objects of the sample data and analysis results were created using the Seurat (v. 4.3.0) R package (https://github.com/scOpenLab/spatial_analysis/tree/main) Submitted via NFDI4BIOIMAGE