NeuroPack: Advanced Learnable Packaging Toolkit Towards Ultra-High Compression Ratio for Volumetric Neuroimaging
- 1Shanghai Artificial Intelligence Laboratory
Abstract
High-throughput 3D microscopic imaging generates brain-volume datasets at terabyte to petabyte scale, posing significant storage challenges for large-scale neuroimaging data. In this study, we introduce NeuroPack, a hyperprior-driven 3D deep-learning-based packaging toolkit specifically designed for compressing 3D microscopic volumes at the mesoscale. NeuroPack encodes a compact hyperprior stream that conditions its entropy model on local voxel-wise variance, adaptively allocating bits to preserve mesoscale anatomical details for an optimal rate-distortion balance. To ensure reproducible deployment across heterogeneous microscopy datasets and computing environments, NeuroPack implements model integerization for the hyperprior side-information pathway, whose hyper-decoder predicts the Gaussian mean and scale parameters used by the arithmetic coder. By replacing platform-dependent floating-point hyper-decoding with deterministic fixed-point integer convolutions, NeuroPack produces bit-exact entropy models across CPUs, GPUs, local workstations, and cloud servers, thereby preventing decoding drift caused by floating-point rounding, fused-multiply-add ordering, or library-specific kernels. This integerized design preserves rate-distortion behavior while maintaining efficient encoding and decoding for large anisotropic 3D microscopy volumes. Our method enables high-fidelity microscopy reconstruction at 100- to 1000-fold compression across diverse modalities and species. Quantitatively, NeuroPack delivers the strongest overall rate-distortion performance in both BD-PSNR and BD-MSSSIM, with average relative gains of 73.4% and 50.8%, respectively, over the best competing codec. Importantly, NeuroPack preserves image fidelity required for downstream morphological reconstruction and analysis, as validated by simultaneous human annotation of original and decompressed confocal whole-brain zebrafish microscopy images. Together, NeuroPack provides a learning-based compression framework explicitly tailored for microscopy data, establishing a scalable storage solution for data-intensive neuroimaging and AI-empowered biological discovery.
Related
- Cerebrovascular Structure-Flow Gradients Revealed by Volumetric Ultrasound Localization Microscopy
- Upcycling banana peduncle fibers into mycelium-based composites for sustainable packaging and thermal insulation
- Vault Packaging Dynamics: Cargo Loading Stabilizes MVP Conformation for High-Payload Cytokine Delivery in Cancer Therapy
- A prophage-encoded anti-phage defence system that prevents DNA packaging by targeting the terminase complex
- A Test for Confounding in Coupling of Multimodal Neuroimaging Data
BibTeX
@misc{zhao2026neuropack,
title = {{NeuroPack: Advanced Learnable Packaging Toolkit Towards Ultra-High Compression Ratio for Volumetric Neuroimaging}},
author = {R. Zhao and X. Yang and J. Wu and Q. Zheng and J. Zhang and J.-Q. Guan and W. Cai and J.-L. Du and L. Bai and X.-F. Du and W. Ouyang and G. Lu and C. Song},
howpublished = {bioRxiv},
year = {2026},
month = oct,
doi = {10.64898/2026.10.29.762731},
url = {https://www.biorxiv.org/content/10.64898/2026.10.29.762731v1},
note = {CC BY-NC},
}
RIS
TY - UNPB
TI - NeuroPack: Advanced Learnable Packaging Toolkit Towards Ultra-High Compression Ratio for Volumetric Neuroimaging
AU - R. Zhao
AU - X. Yang
AU - J. Wu
AU - Q. Zheng
AU - J. Zhang
AU - J.-Q. Guan
AU - W. Cai
AU - J.-L. Du
AU - L. Bai
AU - X.-F. Du
AU - W. Ouyang
AU - G. Lu
AU - C. Song
PY - 2026
DA - 2026/10/29
PB - bioRxiv
DO - 10.64898/2026.10.29.762731
UR - https://www.biorxiv.org/content/10.64898/2026.10.29.762731v1
AB - High-throughput 3D microscopic imaging generates brain-volume datasets at terabyte to petabyte scale, posing significant storage challenges for large-scale neuroimaging data. In this study, we introduce NeuroPack, a hyperprior-driven 3D deep-learning-based packaging toolkit specifically designed for compressing 3D microscopic volumes at the mesoscale. NeuroPack encodes a compact hyperprior stream that conditions its entropy model on local voxel-wise variance, adaptively allocating bits to preserve mesoscale anatomical details for an optimal rate-distortion balance. To ensure reproducible deployment across heterogeneous microscopy datasets and computing environments, NeuroPack implements model integerization for the hyperprior side-information pathway, whose hyper-decoder predicts the Gaussian mean and scale parameters used by the arithmetic coder. By replacing platform-dependent floating-point hyper-decoding with deterministic fixed-point integer convolutions, NeuroPack produces bit-exact entropy models across CPUs, GPUs, local workstations, and cloud servers, thereby preventing decoding drift caused by floating-point rounding, fused-multiply-add ordering, or library-specific kernels. This integerized design preserves rate-distortion behavior while maintaining efficient encoding and decoding for large anisotropic 3D microscopy volumes. Our method enables high-fidelity microscopy reconstruction at 100- to 1000-fold compression across diverse modalities and species. Quantitatively, NeuroPack delivers the strongest overall rate-distortion performance in both BD-PSNR and BD-MSSSIM, with average relative gains of 73.4% and 50.8%, respectively, over the best competing codec. Importantly, NeuroPack preserves image fidelity required for downstream morphological reconstruction and analysis, as validated by simultaneous human annotation of original and decompressed confocal whole-brain zebrafish microscopy images. Together, NeuroPack provides a learning-based compression framework explicitly tailored for microscopy data, establishing a scalable storage solution for data-intensive neuroimaging and AI-empowered biological discovery.
ER -