Mindpeak Breast Metastasis Detection
AI-assisted detection and visualization of lymph node metastases in H&E stained breast tissue.
Overview
Mindpeak Breast Metastasis Detection is an AI-based tool developed by Mindpeak GmbH for detecting and visualizing lymph node metastases associated with breast adenocarcinoma in H&E stained tissue. It is intended for use by cancer experts and pathologists seeking to reduce the time and manual effort involved in reviewing breast lymph node slides. The product is designated for Research Use Only and is not intended for use in diagnostic procedures.
The tool is built on Mindpeak's broader platform of automation tools for visual diagnosis, which uses deep learning and explainable AI (xAI) to assist specialists with tasks such as cell and biomarker quantification, freeing them to focus on work requiring human expertise.
Detection and Analysis Capabilities
- Detects and visualizes metastases in H&E stained breast lymph node tissue
- Capable of identifying small metastases associated with breast adenocarcinoma
- Segments tissue into cancerous and benign areas as the primary quantification output
- Uses a proprietary AI model based on convolutional neural networks for segmentation and tumor detection in H&E stained tissue
Workflow
- Whole-slide images are preprocessed by Mindpeak Breast Metastasis Detection
- Cases are triaged based on the preprocessing results
- When a pathologist opens a case, suspicious areas with detected metastases are presented for review
- The pathologist reviews the flagged areas and provides the final diagnosis according to their own assessment
System Requirements and Integration
- Runs on a standard personal working machine
- Minimum hardware requirements: 32-bit processor Intel Core i5 or better; 4 GB RAM
- Compatible with most scanners and microscope cameras
- Designed to integrate into existing laboratory software environments
Mindpeak Breast Metastasis Detection is listed as coming soon for laboratory use. Mindpeak GmbH is headquartered in Hamburg, Germany, and was founded in 2018, with an additional office in Cambridge, MA, United States.


