Limits...
Age-specific MRI brain and head templates for healthy adults from 20 through 89 years of age.

Fillmore PT, Phillips-Meek MC, Richards JE - Front Aging Neurosci (2015)

Bottom Line: The participants included healthy adults from 20 through 89 years of age.It was found that age-appropriate templates provided less biased tissue classification estimates than age-inappropriate reference data and reference data based on young adult templates.This database is available for use by other investigators and clinicians for their MRI studies, as well as other types of neuroimaging and electrophysiological research.

View Article: PubMed Central - PubMed

Affiliation: Department of Communication Sciences and Disorders, University of South Carolina Columbia, SC, USA.

ABSTRACT
This study created and tested a database of adult, age-specific MRI brain and head templates. The participants included healthy adults from 20 through 89 years of age. The templates were done in five-year, 10-year, and multi-year intervals from 20 through 89 years, and consist of average T1W for the head and brain, and segmenting priors for gray matter (GM), white matter (WM), and cerebrospinal fluid (CSF). It was found that age-appropriate templates provided less biased tissue classification estimates than age-inappropriate reference data and reference data based on young adult templates. This database is available for use by other investigators and clinicians for their MRI studies, as well as other types of neuroimaging and electrophysiological research.

No MeSH data available.


Schematic of segmentation routines used on the IBSR test subjects. All subjects had a baseline segmentation with no priors (“Image”), as well as with the MNI152 prior set (“MNI a priori”). Using age-based average MRI templates, segmentation with priors specified both a priori and a posteriori using image-based averages, as well as a posteriori using MNI-based averages. Each of these was conducted using priors from three template age groups: a 20–24 year template, a five-year age-matched template, and a 10-year age-matched template. The eleven segmented volumes resulting from this, and the manual segmented volume, are represented with dashed-line outlines.
© Copyright Policy - open-access
Related In: Results  -  Collection

License
getmorefigures.php?uid=PMC4389545&req=5

Figure 2: Schematic of segmentation routines used on the IBSR test subjects. All subjects had a baseline segmentation with no priors (“Image”), as well as with the MNI152 prior set (“MNI a priori”). Using age-based average MRI templates, segmentation with priors specified both a priori and a posteriori using image-based averages, as well as a posteriori using MNI-based averages. Each of these was conducted using priors from three template age groups: a 20–24 year template, a five-year age-matched template, and a 10-year age-matched template. The eleven segmented volumes resulting from this, and the manual segmented volume, are represented with dashed-line outlines.

Mentions: First we used the manually guided segmentations from the IBSR datasets in order to evaluate the relative fit between our various automated methods and a “gold standard” manually guided segmentation. The goal of this comparison was not primarily an evaluation of the accuracy of these automated methods, but rather to determine the best “proxy standard” to use for our comparisons of age-specific priors, where manual segmentations were unavailable. For this analysis we calculated segmented GM and WM on the IBSR participants MRI volumes for the following eleven procedures: (1) “Image”, calculated from the T1W with no priors; (2) “MNI-a priori”: calculated using the MNI template segmented GM/WM as priors only at the beginning step (FAST—A option); (3) “Image-AVG-a priori” used the “Image” averaged GM/WM volumes as priors only on the beginning step (FAST–a option); (4) “Image-AVG-a posteriori”: used the “Image” averaged GM/WM volumes as priors on the beginning and posteriori steps (“FAST–p”); and (5) “MNI-a posteriori”: Used the “MNI-averaged” GM/WM on the beginning and posteriori steps (“FAST–p”). For the analyses using the averaged GM/WM volumes as segmenting priors, we used the priors from the young adult template (20–24 years), the age-appropriate five-year template, or the age-appropriate 10-year template. The Dice coefficient, which measures degree of overlap (ranging from 0, or no overlap, to 1, total overlap; Dice, 1945) and represents the intersection of two similarly labeled regions divided by the mean volume of the regions, was used to compare the outputs of our various segmentations with each manually segmented IBSR volume. Figure 2 is a schematic showing the eleven segmented volumes that were compared against the manually segmented IBSR volume. A list of these volumes is as follows:


Age-specific MRI brain and head templates for healthy adults from 20 through 89 years of age.

Fillmore PT, Phillips-Meek MC, Richards JE - Front Aging Neurosci (2015)

Schematic of segmentation routines used on the IBSR test subjects. All subjects had a baseline segmentation with no priors (“Image”), as well as with the MNI152 prior set (“MNI a priori”). Using age-based average MRI templates, segmentation with priors specified both a priori and a posteriori using image-based averages, as well as a posteriori using MNI-based averages. Each of these was conducted using priors from three template age groups: a 20–24 year template, a five-year age-matched template, and a 10-year age-matched template. The eleven segmented volumes resulting from this, and the manual segmented volume, are represented with dashed-line outlines.
© Copyright Policy - open-access
Related In: Results  -  Collection

License
Show All Figures
getmorefigures.php?uid=PMC4389545&req=5

Figure 2: Schematic of segmentation routines used on the IBSR test subjects. All subjects had a baseline segmentation with no priors (“Image”), as well as with the MNI152 prior set (“MNI a priori”). Using age-based average MRI templates, segmentation with priors specified both a priori and a posteriori using image-based averages, as well as a posteriori using MNI-based averages. Each of these was conducted using priors from three template age groups: a 20–24 year template, a five-year age-matched template, and a 10-year age-matched template. The eleven segmented volumes resulting from this, and the manual segmented volume, are represented with dashed-line outlines.
Mentions: First we used the manually guided segmentations from the IBSR datasets in order to evaluate the relative fit between our various automated methods and a “gold standard” manually guided segmentation. The goal of this comparison was not primarily an evaluation of the accuracy of these automated methods, but rather to determine the best “proxy standard” to use for our comparisons of age-specific priors, where manual segmentations were unavailable. For this analysis we calculated segmented GM and WM on the IBSR participants MRI volumes for the following eleven procedures: (1) “Image”, calculated from the T1W with no priors; (2) “MNI-a priori”: calculated using the MNI template segmented GM/WM as priors only at the beginning step (FAST—A option); (3) “Image-AVG-a priori” used the “Image” averaged GM/WM volumes as priors only on the beginning step (FAST–a option); (4) “Image-AVG-a posteriori”: used the “Image” averaged GM/WM volumes as priors on the beginning and posteriori steps (“FAST–p”); and (5) “MNI-a posteriori”: Used the “MNI-averaged” GM/WM on the beginning and posteriori steps (“FAST–p”). For the analyses using the averaged GM/WM volumes as segmenting priors, we used the priors from the young adult template (20–24 years), the age-appropriate five-year template, or the age-appropriate 10-year template. The Dice coefficient, which measures degree of overlap (ranging from 0, or no overlap, to 1, total overlap; Dice, 1945) and represents the intersection of two similarly labeled regions divided by the mean volume of the regions, was used to compare the outputs of our various segmentations with each manually segmented IBSR volume. Figure 2 is a schematic showing the eleven segmented volumes that were compared against the manually segmented IBSR volume. A list of these volumes is as follows:

Bottom Line: The participants included healthy adults from 20 through 89 years of age.It was found that age-appropriate templates provided less biased tissue classification estimates than age-inappropriate reference data and reference data based on young adult templates.This database is available for use by other investigators and clinicians for their MRI studies, as well as other types of neuroimaging and electrophysiological research.

View Article: PubMed Central - PubMed

Affiliation: Department of Communication Sciences and Disorders, University of South Carolina Columbia, SC, USA.

ABSTRACT
This study created and tested a database of adult, age-specific MRI brain and head templates. The participants included healthy adults from 20 through 89 years of age. The templates were done in five-year, 10-year, and multi-year intervals from 20 through 89 years, and consist of average T1W for the head and brain, and segmenting priors for gray matter (GM), white matter (WM), and cerebrospinal fluid (CSF). It was found that age-appropriate templates provided less biased tissue classification estimates than age-inappropriate reference data and reference data based on young adult templates. This database is available for use by other investigators and clinicians for their MRI studies, as well as other types of neuroimaging and electrophysiological research.

No MeSH data available.