Seg-CQ500

Intracranial hemorrhages segmentation labels for 51 CT-scans from the CQ500 dataset (http://headctstudy.qure.ai/dataset). Two trained radiologists from the Karolinska Instituted in Stockholm, labeled 51 scans to provide 3D mask of intracranial hemorrhages. We hope our new labels will promote the comparability of hemorrhage segmentation algorithm in the future and help push the field forward. If you use those labels, please cite our paper in Frontiers in Neuroimaging: ``` Spahr A, Ståhle J, Wang C and Kaijser M (2023)<br> Label-efficient deep semantic segmentation of<br> intracranial hemorrhages in CT-scans.<br> Front. Neuroimaging 2:1157565.<br> doi: 10.3389/fnimg.2023.1157565 ```