Breast PR AI for Breast Cancer Sample Analysis

PR IHC digital slide (left) with Breast PR AI markup (right) identifying negative (blue), weak (yellow), moderate (orange), and strong (red) cells. Tumor boundary highlighted with red lines.

Breast PR AI

Developed by Indica Labs, the Breast PR AI solution leverages deep learning that standardizes scoring of progesterone receptor (PR) in resections, excisions, or biopsies from primary invasive breast cancer samples. The algorithm reports quantitative results and markup images, including tumor detection, intensity score, Allred score, and percentage of weak, moderate, and strong positive PR cells.

Activate AI Now

Breast PR AI is For Research Use Only and not intended for clinical diagnostic use. Breast PR AI accessed via the Aperio HALO AP image management system.

PR IHC-stained breast cancer

Zoom image

Step 1: Breast PR IHC Image

Breast PR AI standardizes precision scoring of PR IHC-stained breast cancer resections, excisions and biopsy samples.

The algorithm can be launched by a user directly in the viewer or configured to automatically run for every breast PR in Aperio HALO AP.

Breast PR AI tumor heatmap markup

Zoom image

Step 2: Breast PR IHC Tumor Detection

An image analysis heatmap mask is generated, denoting regions within the sample containing invasive breast cancer.

Prior to this step, image analysis pre-processing excludes artifact and benign epithelial tissue.

The following metrics are calculated as outputs for each slide:

Tumor Area
Other Area
% Tumor Area
Classified Area

Breast PR AI markup

Zoom image

Step 3: PR Cell Classification

Breast PR AI generates an image mask segmenting cells within the tumor region, denoting the following cell classes:

Negative PR Cell: Blue
Weak PR Cell: Yellow
Moderate PR Cell: Orange
Strong PR Cell: Red

Breast PR AI histogram

Zoom image

Results: Histogram

Breast PR AI generates a histogram displaying the total number of cancer cells classified within the following categories:

PR Negative
Weak PR Positive
Moderate PR Positive
Strong PR Positive

Breast PR AI quantitative results

Zoom image

Results: Quantitative

Breast PR AI generates the following metrics as output for the image:

Total Cells
% PR Positive Cells
% PR Cells (Negative, Weak, Moderate, Strong)
Number of PR Positive Cells
Number of PR Cells (Negative, Weak, Moderate, Strong)
Allred: 0-8
Intensity Score: 0, 1, 2, 3

Breast PR AI in Aperio HALO AP Demonstration

Explore the future of pathology with the Breast PR AI app from Indica Labs. Breast PR AI seamlessly integrates within the Aperio HALO AP enterprise digital pathology platform to identify and score PR tumor cells in primary invasive breast cancer samples.

After selecting, view the IHC slide in the viewer and the Breast PR AI results. Results include overlays and quantitative results. The first set of overlays prepare the tissue for analysis by identifying areas of analyzable tissue. The next overlay identifies areas of tumor in the tissue. The final overlay identifies and classifies cells for PR positivity within the tumor area. Finally, quantitative results at a per slide level are displayed in the assay panel to the right.