In this data descriptor, we present the datasets underlying this study, which are stored on the DRYAD data repository, dataset 1–37.8and examples of each dataset to facilitate initial viewing and downloading, Sample Dataset19.

Dataset 1: Volumetric and 3D-Spatial OPT Assessments of BCM Distribution from Pancreatic Compartments (Splenic Lobe (SL), Duodenal Lobe (DL), Gastric Lobe (GL)) in SHD and MLD Diabetic Mice, 1, 2, and 3 weeks post-injection compared to vehicle controls (SHDvCtrl, MLDvCtrl) and untreated controls (Ctrl) with corresponding blood glucose levels and body weights (Table S1)seven.

Dataset 2: OPT analysis of BCM and GLUT2 3D expression intensities in pancreatic splenic lobes of SHD-induced hyperglycemic mice, in which glycemia was restored by islet transplantation (SHD+Tx). Pancreas were removed 28 days after STZ administration and compared to vehicle control and SHD positive control with corresponding blood glucose levels and body weights (Table S2)8.

Dataset 3: High-Resolution Assessments of Islet Morphology Using LSFM from Representative Samples of Dataset 18.

Each data set is subdivided into data records, depending on the image processing pipeline (see Fig. 3). The raw projection views provided (Data Record A, Data Sets 1 and 2, Data Citation 1 and 5) were generated by an in-house near-infrared scanner -OPT16 as *.tiff files. For data record B, 2D tomographic image datasets were processed and reconstructed into tomographic sections (datasets 1 and 2, data citation 2 and 6). Data record B further includes unprocessed LSFM generated sections (data set 3, data citation 9). For the C data record, Z sections of OPT and LSFM imagery were transformed into Imaris files (*.ims) for assessments of the spatial and quantitative characteristics of the BCM distribution (dataset 1, 2 and 3 , citation of data 3, 7 and 10). The resulting quantitative data was extracted from Imaris as Excel sheets (data record D, data citation 4 and 8) including numerical data on islet volumes, staining intensities and islet sphericity, as well as as data on pancreatic lobular anatomy. Taken together, the presented datasets may facilitate the planning, execution and evaluation of a range of research endeavors regarding STZ-induced diabetes in rodents.

Figure 3

Data processing pipeline. Data record A from data set 1 (data record A, data quote 1) and 2 (data record A, data quote 5) was processed using a set of scripts developed in-house post-scan calculations, DFTA (uniform alignment values) and CLAHE (contrast equalizing insulin-stained islets) before reconstruction into tomographic images (the “DSPOPT” image processing package, including including DFTA (“A-value”) and CLAHE setting can be found: https://github.com/ARDISDataset/DSPOPT). Data record B also includes the LSFM z-stacks from data set 3 (data quotes 2, 6, and 9). Tomographic images converted to Imaris native.ims files were analyzed (data record C, 3, 7 and 10). Volumetric and spatial statistics were extracted in Imaris (data record D, data citation 4 and 8).

A schematic image of the organized data tree is shown in Fig. 4. Data records describe datasets of raw, processed images (tomographic reconstructions), endpoint, and quantitative/spatial data of complete BCM distribution in STZ-induced diabetic mice (SHD, MLD, SHD+ Tx) and their healthy controls (C57BL/6) at 1, 2 and 3 weeks after administration, as well as the vehicle controls (SHDvCtrl and MLDvCtrl). Note that the MLD group at the one week time point had no diabetic animals but were still included in the dataset.

Figure 4
number 4

Schematic illustration of the data folder tree. Schematic illustration illustrating the sub-organization of the data sets incorporated into the data descriptor, including AD data records, treatment groups, time points after STZ administration, sample IDs, imaging channels and scan location (LSFM).

Data record A

The raw projection data sets generated by OPT can be found in “Data Record A Raw OPT Projection Views” (Data Record A Raw OPT Projection Views.zip, Data Citation 1 and 5). Each individual scan is supported by a log file in *.txt format including scan parameters such as exposure times or rotation steps. Individual image files are titled to indicate experimental group, age after administration, animal identification, pancreatic lobe (splenic (SL), duodenal (DL), or gastric (GL)), the channel (insulin or anatomy or for dataset 2 GLUT2) and number of rotation steps (1 step = 0.9 degrees of rotation) of the projection image, for example, ‘SHD_2wk_ID3_SL_Insulin_0398.tif’.

Data record B

The data generated by the tomographic reconstruction of the data processed by the OPT (Axis of rotation, DFTA and CLAHE) can be found in ‘Data record B Tomographic images’ (Data record B Tomographic images.zip, Data citation 2 and 6), as well that the raw data generated LSFM z-sections as ‘Data record B LSFM z-stacks’ (Data record B LSFM z-stacks.zip, Data citation 9). Individual image sections from data set 1 and 2 are annotated to indicate experimental group, age after administration, animal identification, pancreatic lobe (SL, DL, GL respectively) , the channel (insulin or anatomy, or GLUT2 for dataset 2) and sequential z-stack number, for example, ‘Ctrl_1wk_ID4_DL_Anatomy_0554.bmp’. Each LSFM section in record B of data set 3 indicates experimental group, age after administration, location of scanned volume (periphery or center of pancreatic splenic lobe), scan ID, the channel (insulin or anatomy) and the z-stack number, for example, ‘MLD_STZ_2wk_SL_Center_2_Insulin_Z0003.ome.tif.

C data record

Reconstructed data converted to *.ims format with 3D isosurfaced volumes can be found in ‘Data record_C_Isosurfaced volume files’ (Data record_C_Isosurfaced volume files.zip, Data citation 3 and 7) and for Dataset 3 Volume files (Data citation 10 ) . The files contain (for datasets 1 and 2) the islets of Langerhans iso-surfaces based on the insulin channel and the lobular anatomy based on the channel anatomy, and the 3D volumes of the aforementioned structures (dataset 3). They are annotated to indicate experimental group, age after administration, animal identification, and pancreatic lobe (splenic (SL), duodenal (DL), or gastric (GL)), e.g., ‘Ctrl_3w_ID4_SL .ims.

Data record D

The resulting quantitative data of the processed 3D OPT volumes was retrieved from Imaris and can be found in ‘Data record D volutric and spatial statistics’ (Data record D volutric and spatial statistics.zip, Data citation 4 and 8). Reconstructed OPT scans produce isotropic voxels, which allows reliable quantification, whereas LSFM scans in general have distortion in the z-axis, just like the Ultramicroscope II used in this study due to sheet generation. of light being two cones overlapping each other rather than two parallel lines. Therefore, quantitative data is provided for OPT data only. Excel sheets based on OPT are extracted raw numerical information. File titles indicate experimental group, post-administration age, animal identification, pancreatic lobe (splenic (SL), duodenal (DL), or gastric (GL)) and channel information, for example, “Ctrl_3w_ID4_DL_anatomy.csv”. The csv files are subdivided into multiple files, each displaying a different setting for individual islands:

  • Volume

  • Area

  • Position

  • Sphericity

  • Intensity (center, min, max, average, median, sum)

  • Dimensional diameter (x,y,z)

  • Homogeneous center of mass

  • Distance to image border

  • Ellipsoid axis

  • Ellipticity (oblate/prolate)

  • Number of triangles

  • Number of vertices

  • Number of voxels

Supplementary Tables 3 and 4 display the metadata for each sample to support the data records and indicate each sample’s provenance and experimental manipulations performed as well as the resulting data outputs and archived records they form.