wassname/PROSTATEx_masks
1
0
Fork 0
mirror of https://github.com/wassname/PROSTATEx_masks.git synced 2026-09-09 11:15:16 +08:00
Code Issues Packages Projects Releases Wiki Activity

Update README.md

Browse Source
This commit is contained in:
Renato Cuocolo
2020-08-03 14:23:29 +02:00
committed by GitHub
parent 4ffe9f9c0f
commit 4fb7ef70cb
1 changed files with 1 additions and 1 deletions
Download Patch File Download Diff File Expand all files Collapse all files
README.md
+1 -1
View File
@@ -5,7 +5,7 @@
Lesion masks for the [PROSTATEx dataset](https://wiki.cancerimagingarchive.net/display/Public/SPIE-AAPM-NCI+PROSTATEx+Challenges).
The dataset comprises prostate MRI exams with PI-RADS score = 2+ lesions. These are classified as clinically significant or not. Only PI-RADS score = 3+ underwent biopsy and are collected in the PROSTATEx 2 dataset, together with the resulting bioptic lesion Gleason Grade.
The original dataset only provided lesion coordinates and an accompanying screenshot for each finding, with several issues due to misplaced coordinates or not clearly identifiable PI-RADS score = 2 lesions. This repository contains the result of a lesion-by-lesion quality check conducted at the Department of Advanced Biomedical Sciences of the University of Naples "Federico II", in the form of lesion masks on axial T2-weighted and ADC images for all correctly identifiable findings in the PROSTATEx/PROSTATEx2 training dataset. As the ground truth was not available for the test sets, the same quality check was not performed on that data.
The original dataset only provided lesion coordinates and an accompanying screenshot for each finding, with several issues due to misplaced coordinates or not clearly identifiable PI-RADS score = 2 lesions. This repository contains the result of a lesion-by-lesion quality check conducted at the [Department of Advanced Biomedical Sciences](http://www.dises.unina.it/en_GB/web/guest/-/768572-dipartimento-di-scienze-biomediche-avanzate) of the [University of Naples "Federico II"](http://www.unina.it/), in the form of lesion masks on axial T2-weighted and ADC images for all correctly identifiable findings in the PROSTATEx/PROSTATEx2 training dataset. As the ground truth was not available for the test sets, the same quality check was not performed on that data.
We encourage the use of this data for radiomics and machine learning investigations in magnetic resonance imaging for prostate cancer, appropriately referencing the repository. Quality checks on the provided masks or other improvements are also welcome as pull requests to the repository.