SCIP: scalable cytometry image processing using Dask in a high performance computing environment
summary
SCIP is an open-source Python tool that runs single-cell bioimage processing (normalization, segmentation, masking, feature extraction) on top of Dask for distributed execution in HPC environments using MPI for communication. It demonstrates scalable, fault-tolerant processing with Dask features (task scheduling, load balancing, data locality) and can run locally or across multiple compute nodes; benchmark comparisons on a Tier‑1 cluster show much faster runtimes than CellProfiler for large cytometry and confocal microscopy datasets, including use of a GPU for cell identification.
AI-written summary · about
related talks
26mdask-image: distributed image processing for large data
01h 23mData Science at Scale with Dask
01h 03mDask: Next Steps in Parallel Python
38mScalable Machine Learning with Dask
01h 26mDask tutorial
01h 56mParallelizing Scientific Python with Dask
01h 17mScaling PyData with Dask
01h 22mUsing Dask for Parallel Computing in Python