In an article in the December 2012 edition of frontiers in neuroscience
Assistant Professor Arno Klein PhD from the Department of Psychiatry
and Behavioral Science at Stony Brook University and Jason Tourville PhD
of Boston University introduced Mindboggle-101, a dataset of 101
manually labeled magnetic resonance images of the human brain which the
authors described as “the largest and most complete set of free,
publicly accessible, manually labeled human brain images.” They also
introduced and made accessible an improved automated brain labeling
protocol — the Desikan-Killany-Tourville protocol — which is intended to
improve the ease, consistency and accuracy of labeling human cortical
regions.
Labeling surface areas of human brains accurately and
consistently is important for research, clinical and educational
endeavors. Mindboggle-101 provides researchers and clinicians with a
common vocabulary to describe research findings. It establishes a
normative database based on images of healthy subjects against which
findings from clinical populations can be compared. The authors
anticipate that their work will aid in the improvement of algorithms
used to segment and label brain images automatically and that it “may
someday aid clinicians in the diagnosis and prediction of treatment
response for neuropsychiatric disorders.”
The article, which is
titled “101 labeled brain images and a consistent human cortical
labeling protocol” is available at the Mindboggle website, along with all data and related software.
