OptraSCAN’s AI & ML-based Image Analysis Solution
For assessment of Nuclear Biomarker Analysis
Multiphase Intensity And Morphology-Based Nuclear Segmentation
Accurate And Detailed Score Generation
The algorithm identifies all nuclei in the blue plane of the image based on the intensity values for the hematoxylin counterstain and cells exhibiting nuclear positivity in the red plane of the image. Distance transform algorithm further applied on the segmented image yields more seed points to separate and segment as many connected cells as possible. An accurate view of nuclear scores mimicking the pathologist’s interpretation are generated.
Image Deconvolution And Enhancement Feature
This feature filters out residual or nonspecific staining in the cytoplasm or background automatically, separates positively stained nuclei from hematoxylin counterstain to generate reproducible and accurate results of true nuclear staining intensity.
Intuitive User Tunable Algorithm Parameters
The user can further train and tune the analysis parameters to determine inclusion of nuclei in analysis based on cell size, area, and staining intensity and evaluate multiple markers in a quantitative and high-throughput manner.
Features Of OptraSCAN’s Nuclear Algorithm
- Computer assisted whole-slide and Regions of Interest assessment, robust and reproducible machine learning technology for accurate quantification of nuclear biomarker expression
- Identification of nuclear (hematoxylin) and positive optical density vectors
- Nuclear segmentation, separation of overlapping nuclei and further filtering using shape, size, roundness, compactness elongation, etc.
- Easy User tunable analysis parameters applicable to the entire image set, identifying nuclei, training and classifying them as positive or negative Generation and quantification of clinically relevant data for each case
- Single slide analysis and batch processing feature with a few clicks and complete walk away automation