Lab Note · updated 2026-06-08

Why Microscopy Needs Verified Outputs

A measurement you cannot reproduce is not a measurement. Why verification is a first-class pipeline stage, not an afterthought.

TopicsReadout · Report · Cell Painting

High-content imaging turns pixels into decisions. If those decisions are not verified, validated, and reproducible, the model is just an opinion.

This is why Percent Replicating is treated as a gate, not a footnote, and why Batch Correction Without Erasing Biology matters before any readout is trusted.

The trust pillar

Every Fovea Lab pipeline ships a QC-aware report — the number and the evidence it holds up.

Bibliography

1Cell Painting assay protocol

Related notes

The QC-Aware Report
A QC-aware report doesn't just present a result — it surfaces the evidence that the result is trustworthy, structured by the four nested levels of quality and ending in an explicit decision. Its job is to make a number un-trustable on sight when it shouldn't be trusted.
From Masks to Morphology: Features & Self-Supervised Embeddings
Once cells are segmented, you turn masks into numbers — either hand-engineered morphology features or learned embeddings. Each choice trades interpretability against the signal it can capture.
Metadata & Provenance: The Run Manifest
A run manifest captures the provenance and acquisition metadata that turn a folder of images into a dataset you can trust, reproduce, and compare across runs.
A Model Isn't the System
A trained model is a small fraction of a working measurement pipeline. Most of the code, and almost all of the risk, lives in the data, glue, configuration, and validation around it — which is why a model that scores well on a benchmark is not a result you can ship.
Evaluating Segmentation Models in Real Microscopy Workflows
A segmenter's benchmark Dice rarely predicts its behavior in a real workflow. Instance counting, merge/split errors, and cross-instrument generalization are the quantities that decide whether the downstream readout is correct — and standard semantic overlap scores are blind to all three.
Foundation Models Are Not Pipelines
A foundation model is a powerful feature extractor, not a deployable measurement pipeline. It still needs ingestion, QC, preprocessing, confounder correction, validation, and provenance around it — and a frozen encoder makes the surrounding system more important, not less.
Ground Truth Is a Design Problem
Ground truth in microscopy is constructed, not given — every label is the output of an annotation protocol, an annotator, and a fusion rule, each with its own bias and variance. Treating it as a fixed answer key is how inter-rater noise and label bias get baked into every score computed against it.
Image Quality Is Not One Number
Fovea treats microscopy quality as four nested levels — image-space, run/sample, readout, and decision — because a metric at one level says nothing about the levels above it. Full-reference scalars like SSIM and PSNR measure only the first, yet teams report as if they measured the third.
Why mAP, Dice, and SSIM Are Not Enough
Dice, IoU, mAP, and SSIM measure overlap or similarity against a reference — not whether the result supports the decision the experiment exists to make. The discipline is problem-aware metric selection: derive the metric from the question, not from what is easy to compute.
How to Prepare Microscopy Data for AI
Most of the work of applying AI to microscopy is data preparation — formatting, normalization, tiling, and label hygiene — and the single most consequential decision is a split that respects the non-independence of microscopy data. Get the split wrong and every downstream metric is leaked, not earned.
From Research Script to Production Pipeline
A notebook that worked once is not a pipeline, and a community model that scored well on its paper's data is not a validated component. Turning either into something deployable is the engineering work of pinning, wrapping, gating, and validating — most of which the original artifact deliberately skipped.
Uncertainty, Drift, and Failure Modes in Bioimage Analysis
Models fail silently when the data drifts away from what they were trained on — and an overconfident, miscalibrated model gives no warning. Trustworthy deployment needs calibrated uncertainty, explicit distribution-shift detection, and a way to flag failure when there is no ground truth to check against.
Validation Without a Ground Truth
Validation asks whether a microscopy pipeline's outputs are biologically true and fit for purpose — but biology rarely supplies a clean answer key. The discipline is choosing metrics that reflect the question, manufacturing ground truth honestly, and conditioning performance on the experiment.
Verification and Validation Are Two Different Questions
Verification asks whether the pipeline was built right — deterministic, reproducible, correct to spec. Validation asks whether it is the right pipeline — outputs that are biologically true and fit for purpose. Microscopy makes both hard, and conflating them is how silently wrong results get shipped.
Measure Where It Matters
Adaptive, uncertainty-driven acquisition treats the microscope as part of the model — spending photons and time only where the image is uncertain or the biology is happening, instead of scanning everything uniformly.
Cellpose vs CellProfiler for Nuclei Segmentation
When a generalist deep model beats a tuned classical pipeline for nuclei — and when it does not.
Percent Replicating
The reproducibility metric that catches a broken assay before it reaches a biologist.
Batch Correction Without Erasing Biology
Sphering and Harmony remove plate effects — push too hard and you remove the signal too.

Local graph

Bibliography

Related notes

The QC-Aware Report

From Masks to Morphology: Features & Self-Supervised Embeddings

Metadata & Provenance: The Run Manifest

A Model Isn't the System

Evaluating Segmentation Models in Real Microscopy Workflows

Foundation Models Are Not Pipelines

Ground Truth Is a Design Problem

Image Quality Is Not One Number

Why mAP, Dice, and SSIM Are Not Enough

How to Prepare Microscopy Data for AI

From Research Script to Production Pipeline

Uncertainty, Drift, and Failure Modes in Bioimage Analysis

Validation Without a Ground Truth

Verification and Validation Are Two Different Questions

Measure Where It Matters

Cellpose vs CellProfiler for Nuclei Segmentation

Percent Replicating

Batch Correction Without Erasing Biology