HALO AI

1. Annotate regions of interest representing different tissue classes (e.g., tumor, stroma, epithelium).

2. Train a HALO AI network using the annotated data.

3. Apply the trained classifier to new images to automatically classify tissue regions.

4. Review and refine the classification results using interactive markups.

1. Select a pre-trained nuclear segmentation network in HALO AI.

2. Annotate nuclei to refine the segmentation for specific tissue types.

3. Train the network using the annotated examples.

4. Apply the trained network to segment nuclei in whole slide images.

5. Quantify the number of nuclei in different regions of interest.

1. Train an AI-powered membrane segmentation network.

2. Annotate cell membranes in IHC-stained images.

3. Apply the trained network to automatically delineate cell membranes.

4. Measure protein expression levels within the segmented cell membranes.

5. Correlate protein expression with clinical outcomes.

1. Select nuclear and membrane segmentation networks.

2. Provide training examples for different cell phenotypes.

3. Train the network to identify phenotypes of interest.

4. Apply the trained phenotyper to whole slide images.

5. Quantify the number of cells in each phenotype.

1. Train an AI-powered quality control network to detect common artifacts.

2. Apply the network to screen newly scanned slides.

3. Identify slides with artifacts and flag them for review.

4. Improve data quality by removing or re-scanning problematic slides.

1. Annotate examples of the rare event of interest.

2. Train a HALO AI classifier to recognize the rare event.

3. Apply the classifier to scan large tissue sections.

4. Review and validate the detected events using interactive markups.

5. Quantify the number of rare events per tissue section.