See it in your slideRequest a demo
Nemo · IHCBuilt on Nemo

Nemo ER-PR

Hormone receptor scoring on IHC

Oestrogen and progesterone receptor scoring on IHC, cell by cell, with intensity.

Request a demoSee the product sheet
Nemo ER-PR output
Nemo ER-PR: raw slide
IHCNemo ER-PR
Nuclear IHC: positive (green) and negative (red) nuclei detected · drag to compare
The problem

Receptor status is read by eye.

The percentage of positive tumour nuclei and their intensity are read by eye. Near the cut-off, the read decides the treatment.

Low-positive cases decide treatment. Cases around the 1–10 % threshold are the ones a count settles. A per-nucleus score with its intensity is reproducible and reviewable.

What changes

Three things you stop doing by hand

  1. 01
    Every nucleus graded

    Each tumour nucleus classified positive or negative with intensity, from a molecular-anchored IHC model.

  2. 02
    Allred without arithmetic

    Same percentage, same slide, every reader.

  3. 03
    Both receptors, one run

    Whole-tumour scoring rather than a representative field.

Allred, assembled

Proportion plus intensity

0–5Proportion scoreshare of positive tumour nuclei
+
0–3Intensity scoreaverage staining strength
=
0–8Allred scorewith % positive and H-score alongside
Outputs

What you get back

Output
What you get
Format
% positive tumour nucleiER and PR separately, with the cut-off useddata
Intensityweak / moderate / strong distributiondata
Allred / [score][scheme to confirm]data
Nucleus mapevery tumour nucleus, positive or negativeoverlay
Reported to standard

ER and PR reported to the ASCO/CAP hormone receptor testing guideline · Reference →

Evidence

Validation

Nuclear detection on 210,419 annotated nuclei.

Cell detection · 1,740 tiles

F1 0.84 on 210,419 nuclei

Precision0.90
Recall0.80
F10.84
0.84ER
0.84PR

Each tile scored against the annotator who labelled it.

Against the annotators

Within the spread between people

Routine annotation10 tiles, all 7 annotators
Nemo vs each annotator0.81
Annotator vs annotator0.83
Experienced annotators20 tiles, all 7 annotators
Nemo vs each annotator0.85
Annotator vs annotator0.88

Tiles labelled by all seven annotators: 3 pathologists, 4 medical students. Nemo scored against each annotator, compared with annotators scored against each other.

Validation scope
  • Public dataset (EndoNuke), benign endometrium: 12 patients, 19 slides
  • Ventana BenchMark Ultra, CONFIRM anti-ER (SP1) and anti-PR (1E2)
  • One scanner, Leica Aperio AT2, at two resolutions
  • 7 annotators: 3 pathologists, 4 medical students
Not yet validated
  • ER/PR positivity scoring: no reference labels in this dataset
  • Cell typing in endometrial stroma: 30% of stromal nuclei called stroma, 92% of epithelium correct
All Nemo validation results →
Also in the Nemo family
Nemo

Nemo HER2

HER2 scoring on IHC

Read more →Nemo

Nemo PD-L1

PD-L1 scoring on IHC

Read more →Nemo

Nemo Ki67

Ki67 scoring on IHC

Read more →

Try Nemo ER-PR

Run it on a slide in the platform, or ask for the product sheet with the full output specification.

See it in your slideRequest a demo