|
Notes:
The data is located
at https://archive.data.jhu.edu/dataset.xhtml?persistentId=doi:10.7281/T1/CZDPSX
The data consists of 35 different subject's non-contrast CT scan,
with 2 subjects having 2 scans but every other subject having one
scan (in tarballs). One subject additionally has 2 scans, but from
the same time point, but with different convolution kernels. Each
scan has at least one reader's manual segmentation of the image to
delineate the mask of the brain areas (including cerebrospinal
fluid (CSF)). The data are from the Minimally-Invasive Surgery plus
rt-PA for Intracerebral hemorrhage Evacuation (MISTIE) Phases II
and III clinical trials the Clot Lysis: Evaluating Accelerated
Resolution of Intraventricular Hemorrhage (CLEAR) Phase III
clinical trial. Each tarball contains a series of NIfTI
(Neuroimaging Informatics Technology Initiative) files. These files
can be read using common tools such as ITK-SNAP (http://www.itksnap.org/pmwiki/pmwiki.php), Mango ( http://ric.uthscsa.edu/mango/), or Papaya (http://ric.uthscsa.edu/mango/papaya.html). Each image has manual segmentation from 2 different readers
which is a binary image delineating the brain and an automated
estimate of the brain by the method proposed in the paper below.
Changes:
|
