cubic: CUDA-accelerated 3D Bioimage Computing

2025-10-15

AI Summary

Current bioimaging analysis tools exhibit significant limitations in scalability, GPU acceleration support, API standardization, and ecosystem integration—particularly for increasingly large-scale, multidimensional (especially 3D) biological microscopy data. To address this, we introduce a device-agnostic Python library for image analysis that automatically dispatches computations across CPU and GPU backends via a unified NumPy-compatible API. The library natively integrates with the SciPy and scikit-image ecosystems and leverages CuPy and RAPIDS cuCIM for CUDA-accelerated execution. It is the first to systematically support GPU-accelerated 3D deconvolution, segmentation, and feature extraction—achieving identical accuracy while accelerating individual operations by factors of several to over an order of magnitude. Additionally, it enables both interactive exploration and high-throughput automated pipelines, substantially improving efficiency, reproducibility, and interoperability in large-scale biological image analysis.

Technology Category

• Computer Vision: Medical and Biological Imaging
• Machine Learning: Bio-inspired Learning
• Intelligent Robots: Multimodal Perception & Sensor Fusion

Application Category

• Search and Retrieval-Augmented AI: Personalized, context-aware and across-device search
• Web Mining and Content Analysis: Mining multimedia, multimodal, multilingual, cross-lingual Web data
• Semantics and Knowledge: Scalable techniques for the creation, curation, publication, maintenance, and consumption of large, Web-based, structured, reusable, knowledge graphs and ontologies

Abstract

Quantitative analysis of multidimensional biological images is useful for understanding complex cellular phenotypes and accelerating advances in biomedical research. As modern microscopy generates ever-larger 2D and 3D datasets, existing computational approaches are increasingly limited by their scalability, efficiency, and integration with modern scientific computing workflows. Existing bioimage analysis tools often lack application programmable interfaces (APIs), do not support graphics processing unit (GPU) acceleration, lack broad 3D image processing capabilities, and/or have poor interoperability for compute-heavy workflows. Here, we introduce cubic, an open-source Python library that addresses these challenges by augmenting widely used SciPy and scikit-image APIs with GPU-accelerated alternatives from CuPy and RAPIDS cuCIM. cubic's API is device-agnostic and dispatches operations to GPU when data reside on the device and otherwise executes on CPU, seamlessly accelerating a broad range of image processing routines. This approach enables GPU acceleration of existing bioimage analysis workflows, from preprocessing to segmentation and feature extraction for 2D and 3D data. We evaluate cubic both by benchmarking individual operations and by reproducing existing deconvolution and segmentation pipelines, achieving substantial speedups while maintaining algorithmic fidelity. These advances establish a robust foundation for scalable, reproducible bioimage analysis that integrates with the broader Python scientific computing ecosystem, including other GPU-accelerated methods, enabling both interactive exploration and automated high-throughput analysis workflows. cubic is openly available at https://github$.$com/alxndrkalinin/cubic

Problem

• Accelerating 3D bioimage analysis with GPU computing
• Overcoming scalability limitations in large microscopy datasets
• Enhancing interoperability of compute-heavy biomedical workflows

Innovation

• GPU-accelerated Python library for 3D bioimages
• Device-agnostic API using CuPy and cuCIM
• Seamlessly accelerates existing SciPy workflows