arxiv_cv 95% Match Research Paper / Tool Paper Medical imaging researchers,Radiologists,Data scientists in healthcare,Developers of medical AI tools,Researchers using PyRadiomics 1 month ago

PyRadiomics-cuda: a GPU-accelerated 3D features extraction from medical images within PyRadiomics

computer-vision › medical-imaging

📄 Abstract

Abstract: PyRadiomics-cuda is a GPU-accelerated extension of the PyRadiomics library, designed to address the computational challenges of extracting three-dimensional shape features from medical images. By offloading key geometric computations to GPU hardware it dramatically reduces processing times for large volumetric datasets. The system maintains full compatibility with the original PyRadiomics API, enabling seamless integration into existing AI workflows without code modifications. This transparent acceleration facilitates efficient, scalable radiomics analysis, supporting rapid feature extraction essential for high-throughput AI pipeline. Tests performed on a typical computational cluster, budget and home devices prove usefulness in all scenarios. PyRadiomics-cuda is implemented in Python and C/CUDA and is freely available under the BSD license at https://github.com/mis-wut/pyradiomics-CUDA Additionally PyRadiomics-cuda test suite is available at https://github.com/mis-wut/pyradiomics-cuda-data-gen. It provides detailed handbook and sample scripts suited for different kinds of workflows plus detailed installation instructions. The dataset used for testing is available at Kaggle https://www.kaggle.com/datasets/sabahesaraki/kidney-tumor-segmentation-challengekits-19

Key Contributions

Introduces PyRadiomics-cuda, a GPU-accelerated extension of PyRadiomics that significantly reduces processing times for 3D feature extraction from medical images. It maintains full API compatibility, enabling seamless integration into existing AI workflows and facilitating efficient, scalable radiomics analysis.

Business Value

Accelerates the development and deployment of AI models in healthcare by enabling faster radiomics feature extraction, crucial for research and clinical applications like cancer diagnosis and treatment planning.

Paper Metadata

Innovation Type

Software/Tool Development

Deployment Feasibility

Very high. It's a drop-in replacement for existing PyRadiomics users, requiring only GPU hardware.

Limitations Addressed

Computational challenges of extracting 3D features from large volumetric medical datasets,Long processing times for radiomics analysis,Need for efficient, scalable radiomics pipelines

Performance Gains

Dramatically reduces processing times for 3D feature extraction compared to the CPU-based PyRadiomics.

View Code on GitHub

Technical Tags

PyRadiomicsGPU Acceleration3D Feature ExtractionMedical ImagesRadiomicsGeometric ComputationsHigh-ThroughputAI WorkflowsPythonC/CUDA

Research Topics

Medical Imaging AnalysisRadiomicsHigh-Performance ComputingMachine Learning InfrastructureComputational Imaging

Methods & Architectures

GPU acceleration (CUDA)Offloading geometric computations to GPUMaintaining PyRadiomics API compatibility

Applications & Tasks

Medical Imaging Radiology Oncology AI in Healthcare Accelerating 3D Feature ExtractionReducing Computational TimeEnabling Scalable Radiomics Analysis Extracting 3D shape features from medical imagesSpeeding up radiomics analysis for large datasetsIntegrating GPU acceleration into existing PyRadiomics workflows

Related Fields

Medical ImagingRadiomicsHigh-Performance ComputingGPU ComputingMachine Learning InfrastructureBioinformatics

Keywords

PyRadiomicsGPU acceleration3D feature extractionmedical imagesradiomicsgeometric computationhigh-throughputAI workflowsPythonCUDAcomputational efficiencymedical AI

Academic Context

mis-wut #Medical Imaging Analysis#Radiomics#High-Performance Computing#Machine Learning Infrastructure#Computational Imaging

Companies & Organizations

Research Institutions

mis-wut

Technology Stack

Frameworks & Libraries

PyRadiomicsC/CUDA

Programming Languages

PythonC++

ML Infrastructure

GPU computing

Commercial Potential

Potential Products

Accelerated radiomics analysis softwareComponents for medical AI platforms

Target Industries

HealthcareMedical ImagingPharmaceuticalsBiotechnologyRadiology Services

Use Case Examples

Rapidly extracting features from large CT/MRI datasets for cancer researchEnabling real-time radiomics analysis in clinical settingsAccelerating the development of AI models for disease diagnosis and prognosis

Competitive Edge

Provides a significant speedup for radiomics feature extraction by leveraging GPU acceleration while maintaining compatibility with the widely-used PyRadiomics library, making it an attractive upgrade.

Resource Requirements

Compute Needs

Requires a CUDA-enabled GPU.

Data Requirements

3D medical image datasets (e.g., DICOM, NIfTI).

Deployment Constraints

Requires compatible GPU hardware and drivers.

Scalability

Highly scalable due to GPU acceleration, enabling processing of large volumetric datasets efficiently.

Regulatory Considerations

Not a medical device itselfbut used in development of medical AI tools which require regulatory approval.

Production Readiness

Maturity Level

Production-ready Tool

Licensing

BSD license

View Full Paper Back to Papers