Abstract:White matter lesions (WML) commonly occur in older brains and are quantifiable on MRI, often used as a biomarker in Aging research. Although algorithms are regularly proposed that identify these lesions from T2‐fluid‐attenuated inversion recovery (FLAIR) sequences, none so far can estimate lesions directly from T1‐weighted images with acceptable accuracy. Since 3D T1 is a polyvalent and higher‐resolution sequence, it could be beneficial to obtain the distribution of WML directly from it. However a serious diff… Show more
“…A visual review of the perivascular segmentations derived from the FreeSurfer software was not satisfactory for the needs of this study. Therefore, deep learning-based software was developed specifically for the present study, using a method similar to a previous process, 19 to identify and segment both the putaminal region and the perivascular space (Figure 2). Since the aim was a quantitative analysis, an objective, well-delineated, and relatively size-standardized putaminal region over which to perform the measurement was favored.…”
Section: Methodsmentioning
confidence: 99%
“…A visual review of the perivascular segmentations derived from the FreeSurfer software was not satisfactory for the needs of this study. Therefore, deep learning-based software was developed specifically for the present study, using a method similar to a previous process, 19 to identify and segment both the putaminal region and the perivascular space (Figure 2).…”
Section: Mri Analysismentioning
confidence: 99%
“…16 JPSC-AD was established to identify environmental and genetic risk factors for dementia, as well as their interactions. [17][18][19] The aim of the present study was to investigate the associations of BP with EPVS volumes and to investigate the interactions between BP and several variables, including dementia, using cross-sectional data from a general Japanese older population.…”
Background Enlarged perivascular spaces (EPVS) of the brain may be involved in dementia, such as Alzheimer’s disease and cerebral small vessel disease (CSVD). Hypertension has been reported to be a risk factor for dementia and CSVD, but the association between blood pressure (BP) and perivascular spaces is still unclear. The aim of this study was to determine the association between BP and EPVS volumes and to examine the interactions of relevant factors. Methods A total of 9296 community-dwelling subjects aged ≥65 years participated in a brain magnetic resonance imaging and health status screening examination. Perivascular volume was measured using a software package based on deep learning that was developed in-house. The associations between BP and EPVS volumes were examined by analysis of covariance and multiple regression analysis. Results Mean EPVS volumes increased significantly with rising systolic and diastolic BP levels ( P for trend = .003, P for trend<.001, respectively). In addition, mean EPVS volumes increased significantly for every 1-mmHg-increment in systolic and diastolic BPs (both P values <.001). These significant associations were still observed in the sensitivity analysis after excluding subjects with dementia. Conclusions The present data suggest that higher systolic and diastolic BP levels are associated with greater EPVS volumes in cognitively normal older people.
“…A visual review of the perivascular segmentations derived from the FreeSurfer software was not satisfactory for the needs of this study. Therefore, deep learning-based software was developed specifically for the present study, using a method similar to a previous process, 19 to identify and segment both the putaminal region and the perivascular space (Figure 2). Since the aim was a quantitative analysis, an objective, well-delineated, and relatively size-standardized putaminal region over which to perform the measurement was favored.…”
Section: Methodsmentioning
confidence: 99%
“…A visual review of the perivascular segmentations derived from the FreeSurfer software was not satisfactory for the needs of this study. Therefore, deep learning-based software was developed specifically for the present study, using a method similar to a previous process, 19 to identify and segment both the putaminal region and the perivascular space (Figure 2).…”
Section: Mri Analysismentioning
confidence: 99%
“…16 JPSC-AD was established to identify environmental and genetic risk factors for dementia, as well as their interactions. [17][18][19] The aim of the present study was to investigate the associations of BP with EPVS volumes and to investigate the interactions between BP and several variables, including dementia, using cross-sectional data from a general Japanese older population.…”
Background Enlarged perivascular spaces (EPVS) of the brain may be involved in dementia, such as Alzheimer’s disease and cerebral small vessel disease (CSVD). Hypertension has been reported to be a risk factor for dementia and CSVD, but the association between blood pressure (BP) and perivascular spaces is still unclear. The aim of this study was to determine the association between BP and EPVS volumes and to examine the interactions of relevant factors. Methods A total of 9296 community-dwelling subjects aged ≥65 years participated in a brain magnetic resonance imaging and health status screening examination. Perivascular volume was measured using a software package based on deep learning that was developed in-house. The associations between BP and EPVS volumes were examined by analysis of covariance and multiple regression analysis. Results Mean EPVS volumes increased significantly with rising systolic and diastolic BP levels ( P for trend = .003, P for trend<.001, respectively). In addition, mean EPVS volumes increased significantly for every 1-mmHg-increment in systolic and diastolic BPs (both P values <.001). These significant associations were still observed in the sensitivity analysis after excluding subjects with dementia. Conclusions The present data suggest that higher systolic and diastolic BP levels are associated with greater EPVS volumes in cognitively normal older people.
“…The periventricular region was defined as a dilated (9 mm) mask of the ventricles, a subcortical region as a ribbon mask of the below gray matter, and an inner gray matter as the basal ganglia and thalamus, with the rest being labeled as deep white matter. From the lesion segmentation and region definition, we could then compute the total lesion volume in each brain region for all acquisitions of every subject 22 . All volume measurements were automatically performed in a computer, and two researchers (YTat and MM) who were blinded from clinical data verified the accuracy.…”
Section: Methodsmentioning
confidence: 99%
“…From the lesion segmentation and region definition, we could then compute the total lesion volume in each brain region for all acquisitions of every subject. 22 All volume measurements were automatically performed in a computer, and two researchers (YTat and MM) who were blinded from clinical data verified the accuracy.…”
Background and Purpose
In terms of the gut‐brain axis, constipation has been considered to be an important factor of neurodegenerative diseases, although the exact mechanism is still controversial. Herein, we aimed to investigate the contribution of constipation to the progression of dementia in a retrospective study.
Methods
Patients of Alzheimer's disease(AD) and amnestic mild cognitive impairment were consecutively screened between January 2015 and December 2020, and those of whom brain MRI and neuropsychological tests were performed twice were enrolled in this study. Participants were classified into with constipation (Cons[+], n = 20) and without constipation (Cons[−], n = 64) groups. Laboratory data at the first visit were used. Regression analysis was performed in MMSE, ADAS‐Cog, and the volumes of hippocampus on MRI‐MPRAGE images and deep white matter lesions (DWMLs) on MRI‐FLAIR images obtained at two different time points.
Results
The main finding was that the Cons[+] group showed 2.7 times faster decline in cognitive impairment compared with the Cons[−] group, that is, the liner coefficients of ADAS‐Cog were 2.3544 points/year in the Cons[+] and 0.8592 points/year in the Cons[−] groups. Ancillary, changes of DWMLs showed significant correlation with the time span (p < 0.01), and the liner coefficients of DWMLs were 24.48 ml/year in the Cons[+] and 14.83 ml/year in the Cons[−] group, although annual rate of hippocampal atrophy was not different between the two groups. Moreover, serum homocysteine level at baseline was significantly higher in the Cons[+] group than Cons[−] group (14.6 ± 6.4 and 11.5 ± 4.2 nmol/ml, respectively: p = 0.03).
Conclusion
There is a significant correlation between constipation and faster progression of AD symptoms along with expansion of DWMLs.
White matter lesions (WML) commonly occur in older brains and are quantifiable on MRI, often used as a biomarker in Aging research. Although algorithms are regularly proposed that identify these lesions from T2‐fluid‐attenuated inversion recovery (FLAIR) sequences, none so far can estimate lesions directly from T1‐weighted images with acceptable accuracy. Since 3D T1 is a polyvalent and higher‐resolution sequence, it could be beneficial to obtain the distribution of WML directly from it. However a serious difficulty, both for algorithms and human, can be found in the ambiguities of brain signal intensity in T1 images. This manuscript shows that a cross‐domain ConvNet (Convolutional Neural Network) approach can help solve this problem. Still, this is non‐trivial, as it would appear to require a large and varied dataset (for robustness) labelled at the same high resolution (for spatial accuracy). Instead, our model was taught from two‐dimensional FLAIR images with a loss function designed to handle the super‐resolution need. And crucially, we leveraged a very large training set for this task, the recently assembled, multi‐sites Japan Prospective Studies Collaboration for Aging and Dementia (JPSC‐AD) cohort. We describe the two‐step procedure that we followed to handle such a large number of imperfectly labeled samples. A large‐scale accuracy evaluation conducted against FreeSurfer 7, and a further visual expert rating revealed that WML segmentation from our ConvNet was consistently better. Finally, we made a directly usable software program based on that trained ConvNet model, available at
https://github.com/bthyreau/deep-T1-WMH
.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.