Dataset Information

SegAE: Unsupervised white matter lesion segmentation from brain MRIs using a CNN autoencoder.

ABSTRACT: White matter hyperintensities (WMHs) of presumed vascular origin are frequently observed in magnetic resonance images (MRIs) of the elderly. Detection and quantification of WMHs is important to help doctors make diagnoses and evaluate prognosis of their elderly patients, and once quantified, these can act as biomarkers in clinical research studies. Manual delineation of WMHs can be both time-consuming and inconsistent, hence, automatic segmentation methods are often preferred. However, fully automatic methods can be challenging to construct due to the variability in lesion load, placement of lesions, and voxel intensities. Several state-of-the-art lesion segmentation methods based on supervised Convolutional Neural Networks (CNNs) have been proposed. These approaches require manually delineated lesions for training the parameters of the network. Here we present a novel approach for WMH segmentation using a CNN trained in an unsupervised manner, by reconstructing multiple MRI sequences as weighted sums of segmentations of WMHs and tissues present in the images. After training, our method can be used to segment new images that are not part of the training set to provide fast and robust segmentation of WMHs in a matter of seconds per subject. Comparisons with state-of-the-art WMH segmentation methods evaluated on ground truth manual labels from two distinct data sets and six different scanners indicate that the proposed method works well at generating accurate WMH segmentations without the need for manual delineations.

SUBMITTER: Atlason HE

PROVIDER: S-EPMC6861597 | biostudies-literature | 2019

REPOSITORIES: biostudies-literature

ACCESS DATA

Publications

SegAE: Unsupervised white matter lesion segmentation from brain MRIs using a CNN autoencoder.

Atlason Hans E HE Love Askell A Sigurdsson Sigurdur S Gudnason Vilmundur V Ellingsen Lotta M LM

NeuroImage. Clinical 20191109

White matter hyperintensities (WMHs) of presumed vascular origin are frequently observed in magnetic resonance images (MRIs) of the elderly. Detection and quantification of WMHs is important to help doctors make diagnoses and evaluate prognosis of their elderly patients, and once quantified, these can act as biomarkers in clinical research studies. Manual delineation of WMHs can be both time-consuming and inconsistent, hence, automatic segmentation methods are often preferred. However, fully aut ...[more]

PMID: 31835288

Similar Datasets

Project description:Background and purposeWM lesion segmentation is often performed with the use of subjective rating scales because manual methods are laborious and tedious; however, automated methods are now available. We compared the performance of total lesion volume grading computed by use of an automated WM lesion segmentation algorithm with that of subjective rating scales and expert manual segmentation in a cohort of subjects with type 2 diabetes.Materials and methodsStructural T1 and FLAIR MR imaging data from 50 subjects with diabetes (age, 67.7 ± 7.2 years) and 50 nondiabetic sibling pairs (age, 67.5 ± 9.4 years) were evaluated in an institutional review board-approved study. WM lesion segmentation maps and total lesion volume were generated for each subject by means of the Statistical Parametric Mapping (SPM8) Lesion Segmentation Toolbox. Subjective WM lesion grade was determined by means of a 0-9 rating scale by 2 readers. Ground-truth total lesion volume was determined by means of manual segmentation by experienced readers. Correlation analyses compared manual segmentation total lesion volume with automated and subjective evaluation methods.ResultsCorrelation between average lesion segmentation and ground-truth total lesion volume was 0.84. Maximum correlation between the Lesion Segmentation Toolbox and ground-truth total lesion volume (ρ = 0.87) occurred at the segmentation threshold of k = 0.25, whereas maximum correlation between subjective lesion segmentation and the Lesion Segmentation Toolbox (ρ = 0.73) occurred at k = 0.15. The difference between the 2 correlation estimates with ground-truth was not statistically significant. The lower segmentation threshold (0.15 versus 0.25) suggests that subjective raters overestimate WM lesion burden.ConclusionsWe validate the Lesion Segmentation Toolbox for determining total lesion volume in diabetes-enriched populations and compare it with a common subjective WM lesion rating scale. The Lesion Segmentation Toolbox is a readily available substitute for subjective WM lesion scoring in studies of diabetes and other populations with changes of leukoaraiosis.

Project description:BACKGROUND AND PURPOSE:White matter lesions are 1 age-related manifestation of cerebrovascular disease, but subthreshold abnormalities have been identified in nonlesional WM. We hypothesized that structural and physiologic MR imaging findings of early cerebrovascular disease can be measured in middle-aged subjects in tissue adjacent to WM lesions, termed "penumbra." MATERIALS AND METHODS:WM lesions were defined using automated segmentation in 463 subjects, 43-56 years of age, from the Coronary Artery Risk Development in Young Adults (CARDIA) longitudinal observational cohort study. We described 0- to 2-mm and 2- to 4-mm-thick spatially defined penumbral WM tissue ROIs as rings surrounding WM lesions. The remaining WM was defined as distant normal-appearing WM. Mean signal intensities were measured for FLAIR, T1-, and T2-weighted images, and from fractional anisotropy, mean diffusivity, CBF, and vascular reactivity maps. Group comparisons were made using Kruskal-Wallis and pair-wise t tests. RESULTS:Lesion volumes averaged 0.738 ± 0.842 cm3 (range, 0.005-7.27 cm3). Mean signal intensity for FLAIR, T2, and mean diffusivity was increased, while T1, fractional anisotropy, and CBF were decreased in white matter lesions versus distant normal-appearing WM, with penumbral tissues showing graded intermediate values (corrected P < .001 for all group/parameter comparisons). Vascular reactivity was significantly elevated in white matter lesions and penumbral tissue compared with distant normal-appearing white matter (corrected P ? .001). CONCLUSIONS:Even in relatively healthy 43- to 56-year-old subjects with small white matter lesion burden, structural and functional MR imaging in penumbral tissue reveals significant signal abnormalities versus white matter lesions and other normal WM. Findings suggest that the onset of WM injury starts by middle age and involves substantially more tissue than evident from focal white matter lesions visualized on structural imaging.

Dataset Information

SegAE: Unsupervised white matter lesion segmentation from brain MRIs using a CNN autoencoder.

Publications

SegAE: Unsupervised white matter lesion segmentation from brain MRIs using a CNN autoencoder.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets