Mindpeak Breast PR

When a compatible PR image is opened, analysis runs automatically across the entire whole slide image—no region drawing, no manual counting. The algorithm locates the invasive tumor, classifies every tumor cell nucleus as PR-positive or PR-negative, and returns a color-coded overlay within seconds. Because coverage is complete rather than a hand-picked field, the output does not depend on where a researcher happens to look, which is one of the best documented sources of scoring variability. At low magnification, the overlay shows immediately how PR expression is distributed across the slide and how much scoreable tumor it contains. Because the score is derived from the same cells displayed in the overlay, the overlay also provides a clear audit trail. Every tissue fragment on the slide is included in the analysis, including the small or fragmented cores that are easiest to overlook on a manual read.

Region-level interrogation of the whole slide analysis. The 0-click analysis covers the whole slide, but the analysis is not locked to it. Using the standard annotation tools, subregions can be drawn on top of the completed analysis and scored individually, each returning its own positivity score and its own positive and negative cell counts. This supports two capabilities that whole slide scoring alone cannot provide. Tissue can be excluded from consideration, such as a second core, a poorly fixed fragment, crush artifact, or an area the researcher judges unrepresentative. PR expression can also be compared between regions, making intratumoral heterogeneity measurable rather than impressionistic: per core, per tumor focus, or across any regions of interest the study design calls for. Regional findings are presented alongside the whole slide assessment rather than replacing it, ensuring both views remain available when the analysis is complete.

The algorithm returns an PR-positive score when 1% or more of tumor cell nuclei are stained, which is the same as the ER algorithm. PR expression is frequently more heterogeneous than ER within the same tumor, so the percentage has to be estimated across patchy staining, and that is precisely where manual estimation is least reproducible. Mindpeak Breast PR replaces the estimate with an actual count: every tumor cell nucleus on the slide is classified and totaled, so the percentage is computed rather than judged by eye. Difficulty staining interpretations are therefore resolved on the same basis every time, regardless of the researcher, and the count behind the percentage remains visible for review rather than disappearing into a single score.

Each tumor cell nucleus on the slide is delineated and classified according to its PR staining, and the positivity score is reported as the proportion of positive nuclei within the total tumor nucleus population, providing a measured count rather than a visual estimate. The full classification is rendered over the tissue, allowing researchers to see exactly which nuclei produced the percentage score and to assess complex samples based on visible evidence. Because the count is exhaustive rather than sampled, the same slide is assessed on the same basis every time. When a output requires refinement or part of the tissue needs to be excluded, the annotation workflow described above provides that control.

A score with the numbers behind it. Outputs are returned as a positivity score together with the positive and negative tumor cell counts used to generate it, providing visibility into the basis of the output. Individual markup classes can be shown or hidden while reviewing, allowing the researcher to inspect positive or negative cell populations independently before accepting the result. The application was evaluated across four whole slide scanners, and across three PR staining systems with results shown to be agnostic to scanner, staining system, reagent, and operator.