Example data for the paper "cubic: CUDA-Accelerated 3D Bioimage Computing"

Example image data accompanying the cubic Python library (CUDA-accelerated 3D bioimage computing).

Used by the examples/notebooks/deconvolution_iterations_3d.ipynb notebook to demonstrate how PSNR, SSIM, FSC, and DCR can be used to track the progress of Richardson-Lucy 3D deconvolution and inform iteration-count selection.

Files

• astr_vpa_hoechst.tif — A single 3D stack of Hoechst-stained astrocyte nuclei (16-bit). Acquired on a Yokogawa CQ1 confocal microscope. Voxel size: 0.1625 μm (xy) × 0.3 μm (z).
• astr_vpa_hoechst_psf_na095_cropped.tif — Theoretical 3D point spread function (PSF) modeled with the Richards & Wolf algorithm from the PSFGenerator ImageJ/Fiji plugin (NA = 0.95). Center-cropped to 30×210×210 (captures ≥ 99.5% of energy); the Richardson-Lucy deconvolution implementation zero-pads it to the image size internally.

Citation

If you use these data, please cite the cubic paper:

A. A. Kalinin, A. E. Carpenter, S. Singh, and M. J. O’Meara, “Cubic: CUDA-Accelerated 3D Bioimage Computing,” in 2025 IEEE/CVF International Conference on Computer Vision Workshops (ICCVW), Honolulu, HI, USA, Oct. 2025, pp. 5831–5840, doi: 10.1109/ICCVW69036.2025.00608.

BibTeX:

@inproceedings{kalinin2025cubic,
  author    = {Kalinin, Alexandr A. and Carpenter, Anne E. and Singh, Shantanu and O'Meara, Matthew J.},
  title     = {Cubic: {CUDA-Accelerated} {3D} {Bioimage} {Computing}},
  booktitle = {2025 IEEE/CVF International Conference on Computer Vision Workshops (ICCVW)},
  year      = {2025},
  pages     = {5831--5840},
  address   = {Honolulu, HI, USA},
  publisher = {IEEE},
  doi       = {10.1109/ICCVW69036.2025.00608}
}

Open access: CVF proceedings PDF. Preprint: arXiv:2510.14143.