Imaging Pipeline and Analysis Build

Who it is for

Anyone who generates image or cell data and needs one assay’s analysis to be reproducible by someone other than the person who wrote it. Drug developers, CROs, imaging and instrument vendors, academic cores and tool companies buy this for the same reason: the analysis lives in one person’s hands and one vendor’s interface.

Scope

One assay, end to end, in open tools: segmentation, feature extraction, feature selection, multivariate analysis and clustering, controls and normalization, statistics, and the figure-generating code. Built in Python and CellProfiler so you can rerun it without buying a license.

Inputs

A representative image or recording set, the acquisition settings, the current pipeline or vendor settings, and the biological question the readout has to answer.

Deliverable

• a documented, rerunnable notebook you own outright
• a written methods section, at the level a journal or a reviewer expects, describing what the pipeline does and why
• a validation summary showing the pipeline reproduces the reference result on your sample
• a handover session with the person who will run it

Timeline

Three to five weeks, depending on how much of the sample has to be re-acquired.

What is excluded

Acquiring new images, instrument service, and production deployment inside your informatics stack. Models, classifiers and analysis code built on your data are yours; my pre-existing frameworks and methods remain mine and are licensed to you.

Referred out

Immunohistochemistry (IHC); live-animal imaging; electron microscopy. I refer without a fee.