Limits...
Quantifying brain tissue volume in multiple sclerosis with automated lesion segmentation and filling.

Valverde S, Oliver A, Roura E, Pareto D, Vilanova JC, Ramió-Torrentà L, Sastre-Garriga J, Montalban X, Rovira À, Lladó X - Neuroimage Clin (2015)

Bottom Line: Then, images processed following each of the pipelines were segmented into gray matter (GM) and white matter (WM) using SPM8, and compared with the same images where expert lesion annotations were filled before segmentation.However, the % of error was significantly lower when automatically estimated lesions were filled and not masked before segmentation.These results are relevant and suggest that LST and SLS toolboxes allow the performance of accurate brain tissue volume measurements without any kind of manual intervention, which can be convenient not only in terms of time and economic costs, but also to avoid the inherent intra/inter variability between manual annotations.

View Article: PubMed Central - PubMed

Affiliation: Dept. of Computer Architecture and Technology, University of Girona, Spain.

ABSTRACT
Lesion filling has been successfully applied to reduce the effect of hypo-intense T1-w Multiple Sclerosis (MS) lesions on automatic brain tissue segmentation. However, a study of fully automated pipelines incorporating lesion segmentation and lesion filling on tissue volume analysis has not yet been performed. Here, we analyzed the % of error introduced by automating the lesion segmentation and filling processes in the tissue segmentation of 70 clinically isolated syndrome patient images. First of all, images were processed using the LST and SLS toolkits with different pipeline combinations that differed in either automated or manual lesion segmentation, and lesion filling or masking out lesions. Then, images processed following each of the pipelines were segmented into gray matter (GM) and white matter (WM) using SPM8, and compared with the same images where expert lesion annotations were filled before segmentation. Our results showed that fully automated lesion segmentation and filling pipelines reduced significantly the % of error in GM and WM volume on images of MS patients, and performed similarly to the images where expert lesion annotations were masked before segmentation. In all the pipelines, the amount of misclassified lesion voxels was the main cause in the observed error in GM and WM volume. However, the % of error was significantly lower when automatically estimated lesions were filled and not masked before segmentation. These results are relevant and suggest that LST and SLS toolboxes allow the performance of accurate brain tissue volume measurements without any kind of manual intervention, which can be convenient not only in terms of time and economic costs, but also to avoid the inherent intra/inter variability between manual annotations.

No MeSH data available.


Related in: MedlinePlus

% of absolute error in total GM and WM volume between segmented images where the annotated lesion masks were refilled before tissue segmentation (Expert filled) and the same images processed following the Original, Expert masked, SLS/LST masked, and SLS/LST filled pipelines. Results for the SLS toolbox are shown in the top row for GM (A) and WM (B), and for the LST toolbox in the bottom row for GM (C) and WM (D). The Δ symbol depicts the mean % difference in total GM/WM tissue for each pipeline. Horizontal lines show significant differences between evaluated pipelines with *p < 0.05, **p < 0.01, ***p < 0.001.
© Copyright Policy - CC BY-NC-ND
Related In: Results  -  Collection

License
getmorefigures.php?uid=PMC4644250&req=5

f0005: % of absolute error in total GM and WM volume between segmented images where the annotated lesion masks were refilled before tissue segmentation (Expert filled) and the same images processed following the Original, Expert masked, SLS/LST masked, and SLS/LST filled pipelines. Results for the SLS toolbox are shown in the top row for GM (A) and WM (B), and for the LST toolbox in the bottom row for GM (C) and WM (D). The Δ symbol depicts the mean % difference in total GM/WM tissue for each pipeline. Horizontal lines show significant differences between evaluated pipelines with *p < 0.05, **p < 0.01, ***p < 0.001.

Mentions: First, we analyzed the differences in total tissue volume between the images processed following each of the SLS pipelines and the images where expert lesion masks had been filled with the SLF method before tissue segmentation. Automated lesion segmentation and filling reduced significantly the % of error in total GM (p < 0.032) on the images processed with the fully automated SLS filled pipeline when compared with the same images segmented including lesions (Original pipeline) (see Fig. 1A). Similarly, the % differences in total WM were also significantly lower on the Expert masked (p < 0.040) and SLS filled (p < 0.002) pipelines when compared with the Original images (see Fig. 1B). Differences in total GM and WM between the SLS masked and SLS filled pipelines were not statistically different.


Quantifying brain tissue volume in multiple sclerosis with automated lesion segmentation and filling.

Valverde S, Oliver A, Roura E, Pareto D, Vilanova JC, Ramió-Torrentà L, Sastre-Garriga J, Montalban X, Rovira À, Lladó X - Neuroimage Clin (2015)

% of absolute error in total GM and WM volume between segmented images where the annotated lesion masks were refilled before tissue segmentation (Expert filled) and the same images processed following the Original, Expert masked, SLS/LST masked, and SLS/LST filled pipelines. Results for the SLS toolbox are shown in the top row for GM (A) and WM (B), and for the LST toolbox in the bottom row for GM (C) and WM (D). The Δ symbol depicts the mean % difference in total GM/WM tissue for each pipeline. Horizontal lines show significant differences between evaluated pipelines with *p < 0.05, **p < 0.01, ***p < 0.001.
© Copyright Policy - CC BY-NC-ND
Related In: Results  -  Collection

License
Show All Figures
getmorefigures.php?uid=PMC4644250&req=5

f0005: % of absolute error in total GM and WM volume between segmented images where the annotated lesion masks were refilled before tissue segmentation (Expert filled) and the same images processed following the Original, Expert masked, SLS/LST masked, and SLS/LST filled pipelines. Results for the SLS toolbox are shown in the top row for GM (A) and WM (B), and for the LST toolbox in the bottom row for GM (C) and WM (D). The Δ symbol depicts the mean % difference in total GM/WM tissue for each pipeline. Horizontal lines show significant differences between evaluated pipelines with *p < 0.05, **p < 0.01, ***p < 0.001.
Mentions: First, we analyzed the differences in total tissue volume between the images processed following each of the SLS pipelines and the images where expert lesion masks had been filled with the SLF method before tissue segmentation. Automated lesion segmentation and filling reduced significantly the % of error in total GM (p < 0.032) on the images processed with the fully automated SLS filled pipeline when compared with the same images segmented including lesions (Original pipeline) (see Fig. 1A). Similarly, the % differences in total WM were also significantly lower on the Expert masked (p < 0.040) and SLS filled (p < 0.002) pipelines when compared with the Original images (see Fig. 1B). Differences in total GM and WM between the SLS masked and SLS filled pipelines were not statistically different.

Bottom Line: Then, images processed following each of the pipelines were segmented into gray matter (GM) and white matter (WM) using SPM8, and compared with the same images where expert lesion annotations were filled before segmentation.However, the % of error was significantly lower when automatically estimated lesions were filled and not masked before segmentation.These results are relevant and suggest that LST and SLS toolboxes allow the performance of accurate brain tissue volume measurements without any kind of manual intervention, which can be convenient not only in terms of time and economic costs, but also to avoid the inherent intra/inter variability between manual annotations.

View Article: PubMed Central - PubMed

Affiliation: Dept. of Computer Architecture and Technology, University of Girona, Spain.

ABSTRACT
Lesion filling has been successfully applied to reduce the effect of hypo-intense T1-w Multiple Sclerosis (MS) lesions on automatic brain tissue segmentation. However, a study of fully automated pipelines incorporating lesion segmentation and lesion filling on tissue volume analysis has not yet been performed. Here, we analyzed the % of error introduced by automating the lesion segmentation and filling processes in the tissue segmentation of 70 clinically isolated syndrome patient images. First of all, images were processed using the LST and SLS toolkits with different pipeline combinations that differed in either automated or manual lesion segmentation, and lesion filling or masking out lesions. Then, images processed following each of the pipelines were segmented into gray matter (GM) and white matter (WM) using SPM8, and compared with the same images where expert lesion annotations were filled before segmentation. Our results showed that fully automated lesion segmentation and filling pipelines reduced significantly the % of error in GM and WM volume on images of MS patients, and performed similarly to the images where expert lesion annotations were masked before segmentation. In all the pipelines, the amount of misclassified lesion voxels was the main cause in the observed error in GM and WM volume. However, the % of error was significantly lower when automatically estimated lesions were filled and not masked before segmentation. These results are relevant and suggest that LST and SLS toolboxes allow the performance of accurate brain tissue volume measurements without any kind of manual intervention, which can be convenient not only in terms of time and economic costs, but also to avoid the inherent intra/inter variability between manual annotations.

No MeSH data available.


Related in: MedlinePlus