Integrating Pathologist input in development and validation of a Deep Learning Tissue Classifier App

Note: Permission has been granted for the use of this material
Study Background
Integrating Pathologist input in development and validation of a Deep Learning Tissue Classifier App
This case study focuses on the development & validation of a ‘Deep Learning Tumor / Stroma Classifier App’ to be applied to the tumor microenvironment (TME) present per sample.
Ensuring image analysis apps are validated for accuracy is key to their performance and the generation of robust downstream analysis data.
Staining
The quantification of CD8 (red) / Ki67 (brown) dual IHC staining within accurately defined tumor and stroma ROI across n=321 pancreatic cancer samples.
Analysis
Image analysis was performed using Visiopharm software.
View the Case Study
Click the thumbnail below to view the case study PDF:
Or watch the video case study presentation:
Benefits of Working with OracleBio for your Clinical Studies:
- Data you can trust – We have thorough quality control procedures in place throughout our workflow to ensure the data we provide is of a high-quality standard
- Work alongside Clinical Project managers – experienced with ensuring data parity between time-separated batches of clinical samples

