The data material consists of digital whole-slide images from patients diagnosed with primary papillary urothelial carcinoma, collected at the University Hospital of Stavanger, Norway, in the period 2002-2011. The biopsies are formalin-fixed and paraffin-embedded, from which 4 μm slices are cut and stained with Hematoxylin Eosin Saffron (HES).
The prepared tissue samples are scanned at 400x magnification using the Leica SCN400 slide scanner, producing image files in Leica’s SCN file format. The images are stored as a pyramidal tiled image with several down-sampled versions of the base image in the same file to accommodate for rapid zooming. Each level in the file is down-sampled by a factor of 4 from the previous level.Figure 2 shows an example of a pyramidal histological image with three levels. The Vips library27 is capable of extracting the base image as well as the down-sampled versions, making it easy to extract the dataset at each resolution.
Two datasets were collected from the described data material, referred to as the CV dataset and the inference dataset, both are described below.
The prepared tissue samples are scanned at 400x magnification using the Leica SCN400 slide scanner, producing image files in Leica’s SCN file format. The images are stored as a pyramidal tiled image with several down-sampled versions of the base image in the same file to accommodate for rapid zooming. Each level in the file is down-sampled by a factor of 4 from the previous level.
Two datasets were collected from the described data material, referred to as the CV dataset and the inference dataset, both are described below.