SICAPv1 database

DESCRIPTION

SICAPv1 is a public patch-wise database composed by 78 histological Whole Slide Images (WSI) of the prostate. These images were collected by specialists of the Hospital Clínico Universitario de Valencia.

SICAPv1 is composed of 78 WSI: 18 correspond to benign prostate tissue biopsies (negative class) and 60 to pathological prostate tissue biopsies (positive class). This dataset was divided into two subsets, 60 WSI (17 benign and 43 pathological) were used to learn the models and the remaining 18 to test them. The 43 pathological WSI are distributed as follows: 18 WSI diagnosed as grade 3, 15 WSI catalogued as grade 4 and the remaining 10 images were marked as grade 5 by the pathologists.

In order to automatically analyse these gigapixel images, the images weredownsampled from 40× to 10× and divided in patches with a 50% overlap. To test the influence of the patch size, different sizes were selected: 512^2 and 1024^2,resulting on the two different datasets detailed in the following table:

 

 

 BenignGrade 3Grade 4Grade5Malign
WSIS1718151043
512 patch67253805891731142
1024 patch190911318150344

CITE US

@article{Esteban2019,
title = {A new optical density granulometry-based descriptor for the classification of prostate histological images using shallow and deep Gaussian processes},
author = {Ángel E. Esteban, Miguel López-Pérez , Adrián Colomer, María A. Sales, Rafael Molina, Valery Naranjo.},
url = {https://www.sciencedirect.com/science/article/pii/S0169260719303906?via%3Dihub},
issn = {0169-2607},
year = {2019},
date = {2019-09-01},
journal = {Computer Methods and Programs in Biomedicine},
volume = {178},
pages = {303-317},
keywords = {SICAP},
pubstate = {published},
tppubtype = {article}
}

Ángel E. Esteban, Miguel López-Pérez , Adrián Colomer, María A. Sales, Rafael Molina, Valery Naranjo.: A new optical density granulometry-based descriptor for the classification of prostate histological images using shallow and deep Gaussian processes. In: Computer Methods and Programs in Biomedicine, 178 , pp. 303-317, 2019, ISSN: 0169-2607.
dOWNLOAD
DB_SICAPv1
Size: 15GB
Version: 1.0

Agency

Ministerio de Economía, Industria y Competitividad (DPI2016-77869-C2-1-R)   

Years

2017 to 2020

Partners

Quiere más Información