Pituitary Adenoma MRI Segmentation Dataset

Citation Author(s):
Mengqi
Wang
Haijun
Wang
Submitted by:
Mengqi Wang
Last updated:
Tue, 04/18/2023 - 02:52
DOI:
10.21227/66ks-t035
License:
503 Views
Categories:
0
0 ratings - Please login to submit your rating.

Abstract 

Pituitary adenoma (PA) is one of the most common tumors of the central nervous system, accounting for about 10%-25% of all cases. Although it is generally considered to be a benign tumor, the tumor can compress important tissue structures around it and cause corresponding symptoms. And surgical treatment is the preferred treatment for pituitary adenoma. Therefore, preoperative observation on MRI is important to provide anatomical information to plan the resection and predict the prognosis. However, due to the lack of MRI annotation data of pituitary adenoma, there are limited studies on MRI automatic segmentation of pituitary adenoma through deep learning.

Instructions: 

The dataset is retrospectively collected by the First Affiliated Hospital, Sun Yat-sen University, China. All the MRI data were collected with the approval of the local ethics committee. Patients undergo preoperative MR scanning with intact sequences 1 week before the surgery including T1 weighted with contrast (T1C) sequences using a high-field magnetic system (3.0T, Magnetom Verio, Siemens Healthcare, Erlangen, Germany) with a 64-channel Head/Neck coil. It consists of 55 cases of brain MRI images (3D T1 weighted with contrast) with manual annotations of pituitary adenoma (binary masks).