DEEPVIEW: high-resolution microscopy image analysis

DEEPVIEW: A Web based system for visualization and analysis of high resolution microscopy images

DEEPVIEW is a vendor agnostic viewer for digitised tissue samples. The browser based solution enables pathologists to review and annotate the images, while providing a
framework for applying Artificial Intelligence algorithms to analyse the captured images.

The system provides users with an efficient viewer for Whole Slide Images (WSI), that can be accessed from any PACS system, and supports the DICOM Supplement 145.

deepview1

DeepVIEW viewer annotation tool

Current Algorithms

The current Artificial Intelligence algorithms implemented within the app include:

1) Breast Tissue Biopsies analysis:
  • HER2 – classification of HER2 proteins on the surface of cells
  • HE Staining
  • Ki67 – detection of Ki67 hotspot areas
2) Glomerulus:
  • Glomerulus classification and detection in Kidney biopsies
3) Blood Vessel Analysis:
  • Detect a specific type of vessel, and quantify a set of parameters to enable the clinician quantify the vessel
  • Application for detection of Angiogenesis associated with paediatric cancer
deepview2

DeepVIEW Breast Tissue Biopsies artificial intelligence analysis

This project has been funded by the Government of Castilla-La Mancha, Spain, and the European Union by the “Fondo Europeo de Desarrollo Regional”, within the “Innova-Adelante 2019” actions.