TY - JOUR
T1 - Dynamic programming generation of boundaries of local coordinatized submanifolds in the neocortex
T2 - Application to the planum temporale
AU - Ratnanather, J. T.
AU - Barta, Patrick Ernest
AU - Honeycutt, Nancy A
AU - Lee, N.
AU - Morris, H. M.
AU - Dziorny, A. C.
AU - Hurdal, M. K.
AU - Pearlson, Godfrey
AU - Miller, M. I.
N1 - Funding Information:
The work reported here was supported by NIH P41-RR15241 (MIM,JTR), NSF NPACI (MIM,JTR), NIH MH43775, MH60504, MH52886 (PEB,GDP,NAH), MH57180 (MKH), the Susan T. and James H. Bankard, Sr. Undergraduate Research Award for Biomedical Engineering (NL, ACD), NSF Focused Research Group DMS-0101329 (MKH), FSU FYAP-2001 (MKH), and the Mayo Foundation who provided the Analyze software. We thank Keith Doolittle for assistance in algorithmic developments and the anonymous reviewers for their constructive comments.
PY - 2003/9/1
Y1 - 2003/9/1
N2 - Dynamic programming is used to define boundaries of cortical submanifolds with focus on the planum temporale (PT) of the superior temporal gyrus (STG), which has been implicated in a variety of neuropsychiatric disorders. To this end, automated methods are used to generate the PT manifold from 10 high-resolution MRI subvolumes ROI masks encompassing the STG. A procedure to define the subvolume ROI masks from original MRI brain scans is developed. Bayesian segmentation is then used to segment the subvolumes into cerebrospinal fluid, gray matter (GM), and white matter (WM). 3D isocontouring using the intensity value at which there is equal probability of GM and WM is used to reconstruct the triangulated graph representing the STG cortical surface, enabling principal curvature at each point on the graph to be computed. Dynamic programming is used to delineate the PT manifold by tracking principal curves from the retro-insular end of the Heschl's gyrus (HG) to the STG, along the posterior STG up to the start of the ramus and back to the retro-insular end of the HG. A coordinate system is then defined on the PT manifold. The origin is defined by the retro-insular end of the HG and the y-axis passes through the point on the posterior STG where the ramus begins. Automated labeling of GM in the STG is robust with L1 distances between Bayesian and manual segmentation in the range 0.001-0.12 (n = 20). PT reconstruction is also robust with 90% of the vertices of the reconstructed PT within about 1 voxel (n = 20) from semiautomated contours. Finally, the reliability index (based on interrater intraclass correlation) for the surface area derived from repeated reconstructions is 0.96 for the left PT and 0.94 for the right PT, thus demonstrating the robustness of dynamic programming in defining a coordinate system on the PT. It provides a method with potential significance in the study of neuropsychiatric disorders.
AB - Dynamic programming is used to define boundaries of cortical submanifolds with focus on the planum temporale (PT) of the superior temporal gyrus (STG), which has been implicated in a variety of neuropsychiatric disorders. To this end, automated methods are used to generate the PT manifold from 10 high-resolution MRI subvolumes ROI masks encompassing the STG. A procedure to define the subvolume ROI masks from original MRI brain scans is developed. Bayesian segmentation is then used to segment the subvolumes into cerebrospinal fluid, gray matter (GM), and white matter (WM). 3D isocontouring using the intensity value at which there is equal probability of GM and WM is used to reconstruct the triangulated graph representing the STG cortical surface, enabling principal curvature at each point on the graph to be computed. Dynamic programming is used to delineate the PT manifold by tracking principal curves from the retro-insular end of the Heschl's gyrus (HG) to the STG, along the posterior STG up to the start of the ramus and back to the retro-insular end of the HG. A coordinate system is then defined on the PT manifold. The origin is defined by the retro-insular end of the HG and the y-axis passes through the point on the posterior STG where the ramus begins. Automated labeling of GM in the STG is robust with L1 distances between Bayesian and manual segmentation in the range 0.001-0.12 (n = 20). PT reconstruction is also robust with 90% of the vertices of the reconstructed PT within about 1 voxel (n = 20) from semiautomated contours. Finally, the reliability index (based on interrater intraclass correlation) for the surface area derived from repeated reconstructions is 0.96 for the left PT and 0.94 for the right PT, thus demonstrating the robustness of dynamic programming in defining a coordinate system on the PT. It provides a method with potential significance in the study of neuropsychiatric disorders.
UR - http://www.scopus.com/inward/record.url?scp=0141869834&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=0141869834&partnerID=8YFLogxK
U2 - 10.1016/S1053-8119(03)00238-6
DO - 10.1016/S1053-8119(03)00238-6
M3 - Article
C2 - 14527596
AN - SCOPUS:0141869834
SN - 1053-8119
VL - 20
SP - 359
EP - 377
JO - NeuroImage
JF - NeuroImage
IS - 1
ER -