Diagnostic performance of deep learning-based automatic white matter hyperintensity segmentation for classification of the Fazekas scale and differentiation of subcortical vascular dementia

Generalizability of trial criteria on amyloid-lowering therapy against Alzheimer's disease to individuals with mild cognitive impairment or early Alzheimer's disease in the general population

Treatment with anti-amyloid-β monoclonal antibodies slowed cognitive decline in recent RCTs in patients with mild cognitive impairment (MCI) and early dementia due to Alzheimer's disease (AD). However, stringent trial eligibility criteria may affect generalisability to clinical practice. We extracted eligibility criteria for trials of aducanumab, lecanemab and donanemab, and applied these to participants with MCI and early clinical AD dementia from the population-based Rotterdam Study. Participants underwent questionnaires, genotyping, brain-MRI, cognitive testing, and cardiovascular assessment. We determined amyloid status using a validated prediction model based on age and APOE-genotype. Of 968 participants (mean age: 75 years, 56% women), 779 had MCI and 189 dementia. Across trials, around 40% of participants would be ineligible because of predicted amyloid negativity. At least one clinical exclusion criterion was present in 76.3% of participants for aducanumab, 75.8% for lecanemab, and 59.8% for donanemab. Common criteria were cardiovascular disease (35.2%), anticoagulant (31.2%), psychotropic or immunological medication use (20.4%), anxiety or depression (15.9%), or lack of social support (15.6%). One-third were ineligible based on brain-MRI findings alone, similar across trials and predominantly due to cerebral small-vessel disease. Combining amyloid, clinical, and imaging criteria, eligibility ranged from 9% (95% CI:7.0-11.1) for aducanumab, 8% (6.2-9.9) lecanemab to 15% (12.4-17.5) for donanemab. Findings from recent RCTs reporting protective effects of monoclonal antibodies against amyloid-β are applicable to less than 15% of community-dwelling individuals with MCI or early AD. These findings underline that evidence for drug efficacy and safety is lacking for the vast majority of patients with MCI/AD in routine clinical practice.

Orthostatic hypotension is involved in cognitive impairment in patients with multiple system atrophy: a multi-center cohort study in China

BACKGROUND

Orthostatic hypotension (OH) is a common symptom of multiple system atrophy (MSA), however, its role in cognitive impairment and the mechanism in these patients remains unclear. This study aims to assess the role of OH on cognitive impairment in MSA patients, as well as to explore the potential association of cerebral autoregulation (CA) and white matter hyperintensities (WMHs) on cognitive impairment.

METHODS

This observational study was conducted in three general hospitals in China from January 2018 to October 2023, with patients at one center followed up for 6 months after enrollment. The primary outcomes included cognitive function assessed using the Mini-Mental State Examination (MMSE) and Montreal cognitive assessment (MoCA). Secondary outcomes included the results of the Head-up tilt test, scores for CA and the extent of WMHs.

RESULTS

The 132 MSA patients included 72 men (54.54%) with a mean age of 61.16 (7.80) years. Among them, 80 patients (60.61%) had orthostatic hypotension, and 48 patients (36.36%) had cognitive impairment. OH plays an important role in cognitive impairment in MSA patients (OR = 0.328,95% CI 0.135-0.797, P = 0.014). Cognitive impairment was associated with impaired CA (OR = 0.088,95% CI 0.012-0.657, P = 0.018) and severe WMHs (OR = 0.030,95% CI 0.002-0.423, P = 0.009), particularly in the presence of OH.

CONCLUSION

OH is associated with cognitive impairment in MSA patients, and cognitive decline is linked to impaired CA and increased WMHs. Future studies are needed to explore the mechanisms underlying cognitive impairment in MSA patients.

Automated brain segmentation and volumetry in dementia diagnostics: a narrative review with emphasis on FreeSurfer

Background

Dementia can be caused by numerous different diseases that present variable clinical courses and reveal multiple patterns of brain atrophy, making its accurate early diagnosis by conventional examinative means challenging. Although highly accurate and powerful, magnetic resonance imaging (MRI) currently plays only a supportive role in dementia diagnosis, largely due to the enormous volume and diversity of data it generates. AI-based software solutions/algorithms that can perform automated segmentation and volumetry analyses of MRI data are being increasingly used to address this issue. Numerous commercial and non-commercial software solutions for automated brain segmentation and volumetry exist, with FreeSurfer being the most frequently used.

Objectives

This Review is an account of the current situation regarding the application of automated brain segmentation and volumetry to dementia diagnosis.

Methods

We performed a PubMed search for “FreeSurfer AND Dementia” and obtained 493 results. Based on these search results, we conducted an in-depth source analysis to identify additional publications, software tools, and methods. Studies were analyzed for design, patient collective, and for statistical evaluation (mathematical methods, correlations).

Results

In the studies identified, the main diseases and cohorts represented were Alzheimer’s disease (n = 276), mild cognitive impairment (n = 157), frontotemporal dementia (n = 34), Parkinson’s disease (n = 29), dementia with Lewy bodies (n = 20), and healthy controls (n = 356). The findings and methods of a selection of the studies identified were summarized and discussed.

Conclusion

Our evaluation showed that, while a large number of studies and software solutions are available, many diseases are underrepresented in terms of their incidence. There is therefore plenty of scope for targeted research.

Impact of white matter hyperintensity volumes estimated by automated methods using deep learning on stroke outcomes in small vessel occlusion stroke

Introduction

Although white matter hyperintensity (WMH) shares similar vascular risk and pathology with small vessel occlusion (SVO) stroke, there were few studies to evaluate the impact of the burden of WMH volume on early and delayed stroke outcomes in SVO stroke.

Materials and methods

Using a multicenter registry database, we enrolled SVO stroke patients between August 2013 and November 2022. The WMH volume was estimated by automated methods using deep learning (VUNO Med-DeepBrain, Seoul, South Korea), which was a commercially available segmentation model. After propensity score matching (PSM), we evaluated the impact of WMH volume on early neurological deterioration (END) and poor functional outcomes at 3-month modified Ranking Scale (mRS), defined as mRS score >2 at 3 months, after an SVO stroke.

Results

Among 1,718 SVO stroke cases, the prevalence of subjects with severe WMH (Fazekas score ≥ 3) was 68.9%. After PSM, END and poor functional outcomes at 3-month mRS (mRS > 2) were higher in the severe WMH group (END: 6.9 vs. 13.5%, p < 0.001; 3-month mRS > 2: 11.4 vs. 24.7%, p < 0.001). The logistic regression analysis using the PSM cohort showed that total WMH volume increased the risk of END [odd ratio [OR], 95% confidence interval [CI]; 1.01, 1.00-1.02, p = 0.048] and 3-month mRS > 2 (OR, 95% CI; 1.02, 1.01-1.03, p < 0.001). Deep WMH was associated with both END and 3-month mRS > 2, but periventricular WMH was associated with 3-month mRS > 2 only.

Conclusion

This study used automated methods using a deep learning segmentation model to assess the impact of WMH burden on outcomes in SVO stroke. Our findings emphasize the significance of WMH burden in SVO stroke prognosis, encouraging tailored interventions for better patient care.

Under pressure: the interplay of hypertension and white matter hyperintensities with cognition in chronic stroke aphasia

While converging research suggests that increased white matter hyperintensity load is associated with poorer cognition, and the presence of hypertension is associated with increased white matter hyperintensity load, the relationship among hypertension, cognition and white matter hyperintensities is not well understood. We sought to determine the effect of white matter hyperintensity burden on the relationship between hypertension and cognition in individuals with post-stroke aphasia, with the hypothesis that white matter hyperintensity load moderates the relationship between history of hypertension and cognitive function. Health history, Fazekas scores for white matter hyperintensities and Wechsler Adult Intelligence Scale Matrix Reasoning subtest scores for 79 people with aphasia collected as part of the Predicting Outcomes of Language Rehabilitation study at the Center for the Study of Aphasia Recovery at the University of South Carolina and the Medical University of South Carolina were analysed retrospectively. We found that participants with a history of hypertension had increased deep white matter hyperintensity severity (P < 0.001), but not periventricular white matter hyperintensity severity (P = 0.116). Moderation analysis revealed that deep white matter hyperintensity load moderates the relationship between high blood pressure and Wechsler Adult Intelligence Scale scores when controlling for age, education, aphasia severity and lesion volume. The interaction is significant, showing that a history of high blood pressure and severe deep white matter hyperintensities together are associated with poorer Matrix Reasoning scores. The overall model explains 41.85% of the overall variation in Matrix Reasoning score in this group of participants. These findings underscore the importance of considering cardiovascular risk factors in aphasia treatment, specifically hypertension and its relationship to brain health in post-stroke cognitive function.

Modulation of the Association Between Corticospinal Tract Damage and Outcome After Stroke by White Matter Hyperintensities

BACKGROUND AND OBJECTIVES

Motor outcomes after stroke relate to corticospinal tract (CST) damage. The brain leverages surviving neural pathways to compensate for CST damage and mediate motor recovery. Thus, concurrent age-related damage from white matter hyperintensities (WMHs) might affect neurologic capacity for recovery after CST injury. The role of WMHs in post-stroke motor outcomes is unclear. In this study, we evaluated whether WMHs modulate the relationship between CST damage and post-stroke motor outcomes.

METHODS

We used data from the multisite ENIGMA Stroke Recovery Working Group with T1 and T2/fluid-attenuated inversion recovery imaging. CST damage was indexed with weighted CST lesion load (CST-LL). WMH volumes were extracted with Freesurfer's SAMSEG. Mixed-effects beta-regression models were fit to test the impact of CST-LL, WMH volume, and their interaction on motor impairment, controlling for age, days after stroke, and stroke volume.

RESULTS

A total of 223 individuals were included. WMH volume related to motor impairment above and beyond CST-LL (β = 0.178, 95% CI 0.025-0.331, p = 0.022). Relationships varied by WMH severity (mild vs moderate-severe). In individuals with mild WMHs, motor impairment related to CST-LL (β = 0.888, 95% CI 0.604-1.172, p < 0.001) with a CST-LL × WMH interaction (β = -0.211, 95% CI -0.340 to -0.026, p = 0.026). In individuals with moderate-severe WMHs, motor impairment related to WMH volume (β = 0.299, 95% CI 0.008-0.590, p = 0.044), but did not significantly relate to CST-LL or a CST-LL × WMH interaction.

DISCUSSION

WMHs relate to motor outcomes after stroke and modify relationships between motor impairment and CST damage. WMH-related damage may be under-recognized in stroke research as a factor contributing to variability in motor outcomes. Our findings emphasize the importance of brain structural reserve in motor outcomes after brain injury.

Artificial intelligence for volumetric measurement of cerebral white matter hyperintensities on thick-slice fluid-attenuated inversion recovery (FLAIR) magnetic resonance images from multiple centers

We aimed to develop a new artificial intelligence software that can automatically extract and measure the volume of white matter hyperintensities (WMHs) in head magnetic resonance imaging (MRI) using only thick-slice fluid-attenuated inversion recovery (FLAIR) sequences from multiple centers. We enrolled 1092 participants in Japan, comprising the thick-slice Private Dataset. Based on 207 randomly selected participants, neuroradiologists annotated WMHs using predefined guidelines. The annotated images of participants were divided into training (n = 138) and test (n = 69) datasets. The WMH segmentation model comprised a U-Net ensemble and was trained using the Private Dataset. Two other models were trained for validation using either both thin- and thick-slice MRI datasets or the thin-slice dataset alone. The voxel-wise Dice similarity coefficient (DSC) was used as the evaluation metric. The model trained using only thick-slice MRI showed a DSC of 0.820 for the test dataset, which is comparable to the accuracy of human readers. The model trained with the additional thin-slice dataset showed only a slightly improved DSC of 0.822. This automatic WMH segmentation model comprising a U-Net ensemble trained on a thick-slice FLAIR MRI dataset is a promising new method. Despite some limitations, this model may be applicable in clinical practice.

Deep learning applications in vascular dementia using neuroimaging

PURPOSE OF REVIEW

Vascular dementia (VaD) is the second common cause of dementia after Alzheimer's disease, and deep learning has emerged as a critical tool in dementia research. The aim of this article is to highlight the current deep learning applications in VaD-related imaging biomarkers and diagnosis.

RECENT FINDINGS

The main deep learning technology applied in VaD using neuroimaging data is convolutional neural networks (CNN). CNN models have been widely used for lesion detection and segmentation, such as white matter hyperintensities (WMH), cerebral microbleeds (CMBs), perivascular spaces (PVS), lacunes, cortical superficial siderosis, and brain atrophy. Applications in VaD subtypes classification also showed excellent results. CNN-based deep learning models have potential for further diagnosis and prognosis of VaD.

SUMMARY

Deep learning neural networks with neuroimaging data in VaD research represent significant promise for advancing early diagnosis and treatment strategies. Ongoing research and collaboration between clinicians, data scientists, and neuroimaging experts are essential to address challenges and unlock the full potential of deep learning in VaD diagnosis and management.

Correlation between lenticulostriate arteries and white matter microstructure changes in patients with cerebral small vessel disease

To explore the correlation between the number of lenticulostriate arteries (LSAs) and the white matter features in cerebral small vessel diseases (CSVD) by 3T magnetic resonance imaging (MRI). Seventy-one patients with diagnoses of CSVD were prospectively enrolled to undergo 3T MRI examination, including high-resolution vascular wall imaging (VWI) and diffusion tensor imaging (DTI). The LSAs were observed and counted on VWI, and the patients were divided into three groups according to the LSA counts. The presence of white matter hyperintensities (WMHs), lacunes, cerebral microbleeds (CMBs), and enlarged perivascular spaces (EPVS) was assessed in each patient, and a composite CSVD score was calculated. Periventricular and deep white matter hyperintensity (PVWMH, DWMH) volume ratios were obtained based on automatic segmentation. Fractional anisotropy (FA) and mean diffusivity (MD) were processed by using tract-based spatial statistics (TBSS) analysis. These parameters were compared among the three groups. Correlations between the LSA counts and white matter features were also analyzed. There were differences in WMHs (P = 0.001), CMBs (P < 0.001), EPVS (P = 0.017), composite CSVD scores (P < 0.001), PVWMH volume ratios (P = 0.001), DWMH volume ratios (P < 0.001), global FA (P = 0.001), and global MD (P = 0.002) among the three groups. There were correlations between the LSA counts and WMHs (r = -0.45, P < 0.001), CMBs (r = -0.44, P < 0.001), EPVS (r = -0.28, P = 0.020), the composite CSVD score (r = -0.52, P < 0.001), DWMH volume ratio (r = -0.47, P < 0.001), PWMH volume ratio (r = -0.34, P = 0.004), global FA (r = 0.36, P = 0.002), and global MD (r = -0.33, P = 0.005). Diabetes mellitus (OR 3.36, 95% CI 1.06-10.63; P = 0.039) and increased DWMH volume ratios (OR 1.04, 95% CI 1.00-1.08; P = 0.048) were independent risk factors for a decrease in LSA counts. TBSS analysis showed differences among the three groups in global FA and MD after adjusting for age and sex (P < 0.05). The LSA counts was associated with white matter microstructure changes in CSVD and has the potential to represent the extent of subcortical microvascular damage in CSVD patients.

Characterization of MRI White Matter Signal Abnormalities in the Pediatric Population

(1) Background and Purpose: The aim of this study was to retrospectively characterize WMSAs in an unselected patient cohort at a large pediatric neuroimaging facility, in order to learn more about the spectrum of the underlying disorders encountered in everyday clinical practice. (2) Materials and Methods: Radiology reports of 5166 consecutive patients with standard brain MRI (2006-2018) were searched for predefined keywords describing WMSAs. A neuroradiology specialist enrolled patients with WMSAs following a structured approach. Imaging characteristics, etiology (autoimmune disorders, non-genetic hypoxic and ischemic insults, traumatic white matter injuries, no final diagnosis due to insufficient clinical information, "non-specific" WMSAs, infectious white matter damage, leukodystrophies, toxic white matter injuries, inborn errors of metabolism, and white matter damage caused by tumor infiltration/cancer-like disease), and age/gender distribution were evaluated. (3) Results: Overall, WMSAs were found in 3.4% of pediatric patients scanned at our and referring hospitals within the ten-year study period. The majority were found in the supratentorial region only (87%) and were non-enhancing (78% of CE-MRI). WMSAs caused by autoimmune disorders formed the largest group (23%), followed by "non-specific" WMSAs (18%), as well as non-genetic hypoxic and ischemic insults (17%). The majority were therefore acquired as opposed to inherited. Etiology-based classification of WMSAs was affected by age but not by gender. In 17% of the study population, a definite diagnosis could not be established due to insufficient clinical information (mostly external radiology consults). (4) Conclusions: An "integrated diagnosis" that combines baseline demographics, including patient age as an important factor, clinical characteristics, and additional diagnostic workup with imaging patterns can be made in the majority of cases.