Automatic segmentation of brain white matter and white matter lesions in normal aging: comparison of five multispectral techniques

Hernández, María del C. Valdés; Gallacher, Peter J.; Bastin, Mark E.; Royle, Natalie A.; Maniega, Susana Muñoz; Deary, Ian J.; Wardlaw, Joanna M.

doi:10.1016/j.mri.2011.09.016

Cited by 24 publications

(31 citation statements)

References 16 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…The quantitative method we used to calculate WML volumes had very high intra- as well inter-observer reliability and performed better than thresholding on one sequence than other multispectral methods [7,16]. This fact, together with the double-revised visual rating, reduces the influence that the variability in the quality of the assessment can produce in the measured correlations, and may explain the higher values we obtained compared to those reported in other studies [8,11].…”

Section: Discussionmentioning

confidence: 99%

“…It then reduces the colour levels of the fused image to 32 clusters using minimum variance quantisation. In direct comparisons, MCMxxxVI performed better than thresholding using one MRI sequence and better than other multispectral methods when using two or more sequences [16]. The T 2 *-weighted and FLAIR sequences were mapped respectively to the red/green colour space [7].…”

Section: Methodsmentioning

confidence: 99%

See 1 more Smart Citation

Close Correlation between Quantitative and Qualitative Assessments of White Matter Lesions

et al. 2012

Self Cite

View full text Add to dashboard Cite

Background: White matter lesions (WML) increase with age and are associated with stroke, cognitive decline and dementia. They can be visually rated or computationally assessed. Methods: We compared WML Fazekas visual rating scores and volumes, determined using a validated multispectral image-fusion technique, in Magnetic Resonance Imaging from 672 participants of the Lothian Birth Cohort 1936 and sought explanations for subjects in whom the correlation (Spearman’s ρ) between the total Fazekas score (summed deep and periventricular ratings, 0–6) and WML volume did not concur (z-score difference >1). Infarcts were identified separately. Results: The median WML Fazekas score was 2 [inter-quartile range (IQR): 2], median WML volume 7.7 ml (IQR: 13.6 ml) and median infarct volume (n = 95) 0.98 ml. Score and volume were highly correlated (Spearman’s ρ = 0.78, p < 0.001). Infarcts did not alter the correlation. Minor discordance occurred in 94/672 (14%) subjects, most with total Fazekas score of 1 (n = 20, WML volume = 4.5–14.8 ml) or 2 (n = 50, WML volume = 0.1–34.4 ml). The main reasons were: subtle WML identified visually but omitted from the volume; prominent ventricular caps but thin body lining giving a periventricular score of 1/2 but large WML volume, and small deep focal lesions which increase the score disproportionally when beginning to coalesce with little change in WML volume. Conclusions: WML rating scores and volumes provide near-equivalent estimates of WML burden, therefore either can be used depending on research circumstances. Even closer agreement could result from improved computational detection of subtle WML and modified visual ratings to differentiate prominent ventricular caps from thin periventricular linings, and small non-coalescent from early coalescent deep WML.

show abstract

Section: Discussionmentioning

confidence: 99%

Section: Methodsmentioning

confidence: 99%

Close Correlation between Quantitative and Qualitative Assessments of White Matter Lesions

et al. 2012

Self Cite

View full text Add to dashboard Cite

show abstract

“…These tools, however, can also be applied to other brain structures or organ systems that undergo changes due to injury, disease, or during the process of aging 14,21,22 . The strategies presented take advantage of software that allows the alignment of cross-sectional and longitudinal MRI sequences to generate 3D volume representations of specific regions or structures of interest.…”

Section: Discussionmentioning

confidence: 99%

“…A common feature of human aging is enlarged lateral ventricles (ventriculomegaly) and associated periventricular edema as observed by MRI and fluid-attenuated inversion recovery MRI (MRI/FLAIR) [8][9][10][11][12][13][14] . To investigate the relationship between ventriculomegaly and the cellular organization of the ventricle lining, postmortem human MRI sequences were matched with histological preparations of lateral ventricle periventricular tissue.…”

Section: Introductionmentioning

confidence: 99%

3D Modeling of the Lateral Ventricles and Histological Characterization of Periventricular Tissue in Humans and Mouse

Acabchuk

Sun

Wolferz

et al. 2015

JoVE

View full text Add to dashboard Cite

The ventricular system carries and circulates cerebral spinal fluid (CSF) and facilitates clearance of solutes and toxins from the brain. The functional units of the ventricles are ciliated epithelial cells termed ependymal cells, which line the ventricles and through ciliary action are capable of generating laminar flow of CSF at the ventricle surface. This monolayer of ependymal cells also provides barrier and filtration functions that promote exchange between brain interstitial fluids (ISF) and circulating CSF. Biochemical changes in the brain are thereby reflected in the composition of the CSF and destruction of the ependyma can disrupt the delicate balance of CSF and ISF exchange. In humans there is a strong correlation between lateral ventricle expansion and aging. Age-associated ventriculomegaly can occur even in the absence of dementia or obstruction of CSF flow. The exact cause and progression of ventriculomegaly is often unknown; however, enlarged ventricles can show regional and, often, extensive loss of ependymal cell coverage with ventricle surface astrogliosis and associated periventricular edema replacing the functional ependymal cell monolayer. Using MRI scans together with postmortem human brain tissue, we describe how to prepare, image and compile 3D renderings of lateral ventricle volumes, calculate lateral ventricle volumes, and characterize periventricular tissue through immunohistochemical analysis of en face lateral ventricle wall tissue preparations. Corresponding analyses of mouse brain tissue are also presented supporting the use of mouse models as a means to evaluate changes to the lateral ventricles and periventricular tissue found in human aging and disease. Together, these protocols allow investigations into the cause and effect of ventriculomegaly and highlight techniques to study ventricular system health and its important barrier and filtration functions within the brain. Video LinkThe video component of this article can be found at

show abstract

“…But conventional algorithms used in normal data mining process are not efficient and robust to yield good results with expected clinical accuracy in multispectral analysis. Recent multispectral brain MRI analysis attempted to resolve this issue by improved feature extraction with statistical and mathematical concepts like probabilistic methods, fuzzy approaches, algebraic techniques, and so forth, [5][6][7].…”

Section: Introductionmentioning

confidence: 99%

Automated Brain Tissue Classification by Multisignal Wavelet Decomposition and Independent Component Analysis

Sindhumol

Kumar²,

Balakrishnan

2013

ISRN Biomedical Imaging

View full text Add to dashboard Cite

Multispectral analysis is a potential approach in simultaneous analysis of brain MRI sequences. However, conventional classification methods often fail to yield consistent accuracy in tissue classification and abnormality extraction. Feature extraction methods like Independent Component Analysis (ICA) have been effectively used in recent studies to improve the results. However, these methods were inefficient in identifying less frequently occurred features like small lesions. A new method, Multisignal Wavelet Independent Component Analysis (MW-ICA), is proposed in this work to resolve this issue. First, we applied a multisignal wavelet analysis on input multispectral data. Then, reconstructed signals from detail coefficients were used in conjunction with original input signals to do ICA. Finally, Fuzzy C-Means (FCM) clustering was performed on generated results for visual and quantitative analysis. Reproducibility and accuracy of the classification results from proposed method were evaluated by synthetic and clinical abnormal data. To ensure the positive effect of the new method in classification, we carried out a detailed comparative analysis of reproduced tissues with those from conventional ICA. Reproduced small abnormalities were observed to give good accuracy/Tanimoto Index values, 98.69%/0.89, in clinical analysis. Experimental results recommend MW-ICA as a promising method for improved brain tissue classification.

show abstract

Automatic segmentation of brain white matter and white matter lesions in normal aging: comparison of five multispectral techniques

Cited by 24 publications

References 16 publications

Close Correlation between Quantitative and Qualitative Assessments of White Matter Lesions

Close Correlation between Quantitative and Qualitative Assessments of White Matter Lesions

3D Modeling of the Lateral Ventricles and Histological Characterization of Periventricular Tissue in Humans and Mouse

Automated Brain Tissue Classification by Multisignal Wavelet Decomposition and Independent Component Analysis

Contact Info

Product

Resources

About