Dong Woo Kang, Chang Uk Lee, and Hyun Kook Lim. Clin Psychopharmacol Neurosci 2017;15:89-99. https://doi.org/10.9758/cpn.2017.15.2.89

Multi-label brain segmentation from brain magnetic resonance imaging (MRI) provides valuable structural information for most neurological analyses. Due to the complexity of the brain segmentation algorithm, it could delay the delivery of neuroimaging findings. Therefore, we introduce Split-Attention U-Net (SAU-Net), a convolutional neural network with skip pathways and a split-attention module that segments brain MRI scans. The proposed architecture employs split-attention blocks, skip pathways with pyramid levels, and evolving normalization layers. For efficient training, we performed pre-training and fine-tuning with the original and manually modified FreeSurfer labels, respectively. This learning strategy enables involvement of heterogeneous neuroimaging data in the training without the need for many manual annotations. Using nine evaluation datasets, we demonstrated that SAU-Net achieved better segmentation accuracy with better reliability that surpasses those of state-of-the-art methods. We believe that SAU-Net has excellent potential due to its robustness to neuroanatomical variability that would enable almost instantaneous access to accurate neuroimaging biomarkers and its swift processing runtime compared to other methods investigated.

Introduction: Mild cognitive impairment (MCI) is a prodromal stage of dementia. The association of body mass index (BMI) and progression to Alzheimer's disease (AD) in MCI subjects according to age, sex, and cognitive intervention remains unknown. We investigated the relationship between BMI and the risk of progression to AD in subjects with MCI, as well as the effect of BMI on progression to AD depending on age, sex, cognitive intervention, and chronic diseases. Methods: Three hundred and eighty-eight MCI subjects were followed for 36.3 ± 18.4 months, prospectively. They underwent neuropsychological testing more than twice during the follow-up period. The MCI subjects were categorized into underweight, normal weight, overweight, and obese subgroups. The associations between baseline BMI and progression to AD over the follow-up period were estimated using Cox proportional hazard regression models. Data were analyzed after stratification by age, sex, cognitive intervention, and chronic diseases. Results: After adjustment for the covariates, the underweight MCI group had a higher risk of progression to AD [hazard ratio (HR): 2.38, 95% confidence interval (CI): 1.17–4.82] relative to the normal weight group. After stratifying by age, sex, cognitive intervention, and chronic diseases, this effect remained significant among females (HR: 3.15, 95% CI: 1.40–7.10), the older elderly ≥75 years old (HR: 3.52, 95% CI: 1.42–8.72), the non-intervention group (HR: 3.06, 95%CI: 1.18–7.91), and the hypertensive group (HR: 4.71, 95% CI: 1.17–18.99). Conclusion: These data indicate that underweight could be a useful marker for identifying individuals at increased risk for AD in MCI subjects. This association is even stronger in females, older elderly subjects, the non-cognitive intervention group, and the hypertensive group.

Objective Alzheimer’s disease (AD) is the most common type of dementia and the prevalence rapidly increased as the elderly population increased worldwide. In the contemporary model of AD, it is regarded as a disease continuum involving preclinical stage to severe dementia. For accurate diagnosis and disease monitoring, objective index reflecting structural change of brain is needed to correctly assess a patient’s severity of neurodegeneration independent from the patient’s clinical symptoms. The main aim of this paper is to develop a random forest (RF) algorithm-based prediction model of AD using structural magnetic resonance imaging (MRI).Methods We evaluated diagnostic accuracy and performance of our RF based prediction model using newly developed brain segmentation method compared with the Freesurfer’s which is a commonly used segmentation software.Results Our RF model showed high diagnostic accuracy for differentiating healthy controls from AD and mild cognitive impairment (MCI) using structural MRI, patient characteristics, and cognitive function (HC vs. AD 93.5%, AUC 0.99; HC vs. MCI 80.8%, AUC 0.88). Moreover, segmentation processing time of our algorithm (<5 minutes) was much shorter than of Freesurfer’s (6–8 hours).Conclusion Our RF model might be an effective automatic brain segmentation tool which can be easily applied in real clinical practice.

Objective: Late-life depression and subjective cognitive decline (SCD) are significant risk factors for dementia. However, studies with a large sample size are needed to clarify their independent and combined risks for subsequent dementia. Methods: This nationwide population-based cohort study included all individuals aged 66 years who participated in the National Screening Program between 2009 and 2013 (N = 939,099). Subjects were followed from the day they underwent screening to the diagnosis of dementia, death, or the last follow-up day (December 31, 2017). Results: Depressive symptom presentation, recent depressive disorder, and SCD independently increased dementia incidence with adjusted hazard ratio (aHR) of 1.286 (95% CI:1.255–1.318), 1.697 (95% CI:1.621–1.776), and 1.748 (95% CI: 689–1.808) respectively. Subjects having both SCD and depression had a higher risk (aHR = 2.466, 95% CI:2.383–2.551) of dementia than having depression (aHR = 1.402, 95% CI:1.364–1.441) or SCD (aHR = 1.748, 95% CI:1.689–1.808) alone. Conclusions: Depressive symptoms, depressive disorder, and SCD are independent risk factors for dementia. Co-occurring depression and SCD have an additive effect on the risk of dementia; thus, early intervention and close follow up are necessary for patients with co-occurring SCD and depression.

There is a growing literature on the impact of ethnicity on brain structure and function. Despite the regional heterogeneity in age-related changes and non-uniformity across brain morphometry measurements in the aging process, paucity of studies investigated the difference in cortical anatomy between the East Asian and Caucasian older adults. The present study aimed to compare cortical anatomy measurements, including cortical thickness, volume and surface area, between cognitively normal East Asian (n = 171) and Caucasian (n = 178) older adults, using surface-based morphometry and vertex-wise group analysis of high-dimensional structural magnetic resonance imaging (MRI) data. The East Asian group showed greater cortical thickness and larger cortical volume in the right superior temporal gyrus, postcentral gyrus, bilateral inferior temporal gyrus, and inferior parietal cortex. The Caucasian group showed thicker and larger cortex in the left transverse temporal cortex, lingual gyrus, right lateral occipital cortex, and precentral gyrus. Additionally, the difference in surface area was discordant with that in cortical thickness. Differences in brain structure between the East Asian and Caucasian might reflect differences in language and information processing, but further studies using standardized methods for assessing racial characteristics are needed. The research results represent a further step towards developing a comprehensive understanding of differences in brain structure between ethnicities of older adults, and this would enrich clinical research on aging and neurodegenerative diseases.