Background The 2018 NIA-AA framework outlines a six-stage continuum from asymptomatic individuals to severe Alzheimer's disease (AD) dementia, but most Chinese research still focuses on dementia or broad diagnostic categories. Objective To map diagnosis, treatment, and care patterns across all six AD clinical stages in China and identify demographic, clinical, treatment and care-related factors associated with disease stage. Methods We conducted a nationwide, open online survey via official media channels targeting patients with clinician-confirmed AD and their caregivers. Data were collected via Questionnaire Star. Descriptive analyses, group comparisons, and ordinal logistic regression were performed to examine factors associated with NIA-AA stage. Results A total of 1116 valid responses were analyzed. Most participants were at Stage 2 or higher, with distribution of 0.4%, 9.1%, 16.0%, 24.8%, 26.6%, and 23.0%, across Stage 1-6. Overall, 64.5% had been diagnosed within five years. Neurology (66.4%) and memory clinics (19.2%) were the most frequently visited departments. Donepezil (52.2%) and Memantine (38.8%) were the most common medications, while 34.5% reported engaging in non-pharmacological interventions. Only 1.9% of patients receiving professional dementia institutional care. In logistic regression, disease duration (OR = 0.724, p = 0.006), stage at first outpatient visit (OR= 1.843, p < 0.001), and Donepezil use (OR = 1.394, p = 0.003) were independently associated with current NIA-AA stage. Conclusions This study provides the first nationwide, real-world description of diagnosis, treatment, and care across all NIA-AA stages in China. The findings highlight the need for improved primary-care screening, expanded memory-clinic access, and structured caregiver support to promote earlier detection and more equitable, stage-appropriate management of AD.

BackgroundSubjective cognitive decline (SCD) is a common early complaint in mild cognitive impairment (MCI). Evidence for the 21-item SCD-Questionnaire (SCD-Q21) to discriminate MCI from normal controls (NCs) is limited.ObjectiveTo investigate the discrimination performance of Chinese SCD-Q21 and compare it with SCD-Q9 for community-based MCI early detection, assess the added value of simple covariates, and determine an optimal SCD-Q21 cut-off.Methods294 NCs and 83 people with MCI were assessed and collected demographic and clinical data. Participants completed SCD-Q21, SCD-Q9, Hamilton Anxiety Scale (HAMA) and Hamilton Depression Scale (HAMD) scale; clinical adjudication used Montreal Cognitive Assessment-Basic, Clinical Dementia Rating, and Activities of Daily Living. Group comparison, logistic regression and ROC analyses were applied. Optimal cut-offs were derived using the Youden index and AUCs were compared using DeLong tests. Within-MCI analyses contrasted screen-positive versus screen-negative subgroups.ResultsTotal SCD-Q21 scores were higher in MCI, although five items [question 1 (Q1), Q2, Q3, Q11, and Q17] did not differ between groups. In multivariable binary logistic regression models, lower education (OR = 0.786), higher body mass index (BMI) (OR = 17.874), and higher SCD-Q21 total scores (OR = 1.114) were independently associated with MCI, whereas SCD-Q9 was not. Standalone AUCs were 0.662 (SCD-Q21) and 0.640 (SCD-Q9). Combing age, sex, education, BMI, and HAMA/HAMD with SCD instrument yielded AUC ∼0.91. SCD-Q21 ≥ 7 gave 69.88% sensitivity and 62.93% specificity. Screen-negative MCI cases showed lower vascular/metabolic comorbidity and lower HAMA/HAMD scores.ConclusionsSCD-Q21 provides independent information but modest stand-alone discrimination. As part of a brief multivariable triage including education, BMI, vascular risk review, and anxiety rating, it supports efficient case-finding in community settings.

Background Striatal dopaminergic deficits, established with iodine 123-2β-carbomethoxy-3β-(4-iodophenyl)-N-(3-fluoropropyl)-nortropan (123I-FP-CIT) SPECT, support the diagnosis of Parkinson disease (PD) or atypical parkinsonian syndrome in clinical uncertainty. The swallow tail sign (STS) at susceptibility-weighted (SW) MRI helps differentiate patients with PD from controls, but its utility in clinically uncertain parkinsonian syndromes remains unclear. Purpose To compare the diagnostic performance of STS absence at SW MRI in diagnosing PD with that of 123I-FP-CIT SPECT in participants with clinically uncertain parkinsonian syndrome. Materials and Methods This prospective, multicenter study included participants with clinically uncertain parkinsonian syndrome between May 2016 and May 2019. Imaging modality specialists independently assessed images from SW MRI and 123I-FP-CIT SPECT while blinded to clinical findings. Diagnostic performance was assessed against diagnosis at clinical follow-up as the reference standard. Performance metrics were compared between modalities and post hoc subgroups split by diagnostic confidence and motor severity using Z tests. Cohen κ was computed for intrarater, interrater, and intermodality reliability. Results A total of 106 participants with clinically uncertain parkinsonian syndrome and diagnostic 3-T SW MRI examinations were included in the study sample (median age, 69 years [IQR, 14 years]; 59 male). STS assessment showed substantial agreement within (κ = 0.69; 95% CI: 0.55, 0.83) and between (κ = 0.70; 95% CI: 0.56, 0.84) raters. STS absence demonstrated a sensitivity of 81% (95% CI: 70, 90) and specificity of 75% (95% CI: 58, 88) in predicting PD. Post hoc analysis demonstrated better diagnostic performance in the subgroup with high (60 of 106 participants) versus low diagnostic confidence, with 91% sensitivity (95% CI: 79, 98; Z = 2.8; P = .005), 87% specificity (95% CI: 60, 98; Z = 1.4; P = .17), and 90% accuracy (95% CI: 79, 96; Z = 3.1; P = .002). 123I-FP-CIT SPECT yielded higher sensitivity than STS (98% [95% CI: 91, 100]; Z = 3.3; P = .001) but only 55% specificity (95% CI: 39, 70; Z = -1.9; P = .06). Conclusion In participants with clinically uncertain parkinsonian syndrome, STS absence on 3-T brain SW MRI scans supported the diagnosis of PD with high accuracy overall and greater accuracy when the diagnostic confidence of STS assessment was high. Clinical trial registration no. NCT03022357 © RSNA, 2025 Supplemental material is available for this article.

Background and Objective: The maturation of ultra-high-field magnetic resonance imaging (MRI) [>= 7 Tesla (7T)] has improved our capability to depict and characterise brain structures efficiently, with better signal-to-noise ratio (SNR) and spatial resolution. We evaluated whether these improvements benefit the clinical detection and management of Parkinson's disease (PD). Methods: We performed a literature search in March 2023 in PubMed (MEDLINE), EMBASE and Google Scholar for articles on "7T MRI" AND " Parkinson*", written in English, published between inception and 1st March, 2023, which we synthesised in narrative form. Key Content and Findings: In deep-brain stimulation (DBS) surgical planning, early studies show that 7T MRI can distinguish anatomical substructures, and that this results in reduced adverse effects. In other areas, while there is strong evidence for improved accuracy and precision of 7T MRI-based measurements for PD, there is limited evidence for meaningful clinical translation. In particular, neuromelanin-iron complex quantification and visualisation in midbrain nuclei is enhanced, enabling depiction of nigrosomes 1-5, improved morphometry and vastly improved radiological assessments; however, studies on the related clinical outcomes, diagnosis, subtyping, differentiation of atypical parkinsonisms, and monitoring of treatment response using 7T MRI are lacking. Moreover, improvements in clinical utility must be great enough to justify the additional costs. Conclusions: Together, current evidence supports feasible future clinical implementation of 7T MRI for PD. Future impacts to clinical decision making for diagnosis, differentiation, and monitoring of progression or treatment response are likely; however, to achieve this, further longitudinal studies using 7T MRI are needed in prodromal, early-stage PD and parkinsonism cohorts focusing on clinical translational potential.

Parkinson's disease is a prevalent neurodegenerative disorder affecting millions of people worldwide. It is caused by degeneration of dopaminergic neurons of the substantia nigra. This has given rise to imaging-based biomarker discovery studies using neuromelanin-sensitive MRI. In this paper, we propose a normative template constructed using a combination of structural and neuromelanin-sensitive MRI that can be used in biomarker studies of Parkinson's disease. In contrast to previous template construction approaches, the proposed combination of images allows to impose further constraints on the registrations used to create the template. Our experiments demonstrate that the proposed template construction pipeline yields a consistent localisation of the substantia nigra and other brainstem structures and that the registrations used for this purpose yield smooth and invertible transformations.

Abstract Background Subjective cognitive decline‐questionnaire 9 (SCD‐Q9) was developed to detect SCD complaints at risk of mild cognitive impairment (MCI). However, our previous findings indicated that its coverage might be insufficient. To test this hypothesis, we recently translated SCD‐Q21. Objective To examine the reliability and validity of this translated SCD‐Q21 and to explore its effectiveness for discriminating MCI from controls. Methods Item analysis was performed to understand its item discrimination and homogeneity. The Cronbach's α and Spearman‐Brown's split‐half coefficients were calculated to test its reliability. The Kaiser–Meyer–Olkin (KMO) value, Bartlett's sphericity test, and exploratory factor analysis (EFA) were used to examine its construct validity. The content validity was evaluated using five‐grade Likert scale. Finally, the SCD‐Q21 scores in MCI and controls were compared. Results The difference of each item between the extreme groups was significant. The Cronbach's α coefficient was .913 and Spearman‐Brown's split‐half coefficient was .894. When performing holding one‐out approach, the Cronbach's α coefficient ranged from .906 to .914. The KMO value was .929 and the difference of Bartlett's Sphericity test was significant. All experts scored 5 points when assessing its content. Finally, a significant difference of score was found between MCI and NC groups. Conclusions The reliability and validity of the SCD‐Q21 are good, which may pave a way for its application in a wider Chinese‐speaking population.

Background Clinical diagnosis and monitoring of Parkinson's disease (PD) remain challenging because of the lack of an established biomarker. Neuromelanin-magnetic resonance imaging (NM-MRI) is an emerging biomarker of nigral depigmentation indexing the loss of melanized neurons but has unknown prospective diagnostic and tracking performance in multicenter settings. Objectives The aim was to investigate the diagnostic accuracy of NM-MRI in early PD in a multiprotocol setting and to determine and compare serial NM-MRI changes in PD and controls. Methods In this longitudinal case-control 3 T MRI study, 148 patients and 97 controls were included from six UK clinical centers, of whom 140 underwent a second scan after 1.5 to 3 years. An automated template-based analysis was applied for subregional substantia nigra NM-MRI contrast and volume assessment. A point estimate of the period of prediagnostic depigmentation was computed. Results All NM metrics performed well to discriminate patients from controls, with receiver operating characteristic showing 85% accuracy for ventral NM contrast and 83% for volume. Generalizability using a priori volume cutoff was good (79% accuracy). Serial MRI demonstrated accelerated NM loss in patients compared to controls. Ventral NM contrast loss was point estimated to start 5 to 6 years before clinical diagnosis. Ventral nigral depigmentation was greater in the most affected side, more severe cases, and nigral NM volume change correlated with change in motor severity. Conclusions We demonstrate that NM-MRI provides clinically useful diagnostic information in early PD across protocols, platforms, and sites. It provides methods and estimated depigmentation rates that highlight the potential to detect preclinical PD and track progression for biomarker-enabled clinical trials. (c) 2022 The Authors. Movement Disorders published by Wiley Periodicals LLC on behalf of International Parkinson and Movement Disorder Society

Objective: Subjective cognitive decline (SCD) complaints as the early manifestation of mild cognitive impairment (MCI) may be harbingers of objective cognitive decline. SCD-questinnaire9 (SCD-Q9) is developed to investigate the early sign for MCI. However, few studies have reported its power for discriminating MCI from healthy controls (HCs). Therefore, this study aims to investigate the discrimination power of SCD-Q9 as a brief screening tool for early detection of SCD in MCI.Methods: 84 HCs and 205 people with MCI were recruited. Their demographic information and scores of SCD-Q9 were compared. A binary logistic regression model was used to analyze the potential affecting factors of MCI, and the Receiver Operating Characteristic analysis was applied to test the discrimination powers of those factors, including SCD-Q9.Results: (1) Single and total scores of SCD-Q9 were all lower in the MCI group than those in the HC group. (2) Ageing, lower education and higher total scores of SCD-Q9 were associated with MCI. (3) Area Under the Curves (AUC) of SCD-Q9 for discriminating MCI from HC group was 0.815 and when integrating age and education, the AUC improved slightly and reached 0.839. Additionally, the sensitivity and specificity were 68.8% and 85.7%, respectively when a cut-off value of 3 was applied. Conclusions: SCD-Q9 may be able to detect the subjective cognitive decline in MCI early, but it may be used together with other screening questionnaires to improve its sensitivity and further verification of its power is needed.

Background Previous reports on APOE ε4 allele distribution in different populations have been inconclusive. The Subjective Cognitive Decline-Questionnaire 9 (SCD-Q9) was developed to identify those at risk of objective cognitive impairment [OCI; including mild cognitive impairment (MCI) and dementia groups), but its association with APOE ε4 and discriminatory powers for SCDwith subtle cognitive decline (SCDs) and OCI in memory clinics are unclear. Objectives To investigate demographic distribution of APOE ε4, its association with SCD-Q9 scores, and its ability to discriminate SCDs and OCI groups from normal control (NC). Methods A total of 632 participants were recruited (NC = 243, SCDs = 298, OCI = 91). APOE ε4 allele distribution and association with SCD-Q9 scores were calculated and the effects on cognitive impairment were analyzed. Receiver operating characteristic (ROC) analysis was applied to identify discriminatory powers for NC, SCDs, and OCI. Results Total APOE ε4 frequency was 13.1%. This did not vary by demography but was higher in patients with OCI. The SCD-Q9 scores were higher in APOE ε4 carriers than non-carriers in the OCI group. The area under the curve (AUC) for discriminating from OCI using APOE ε4 were 0.587 and 0.575, using SCD-Q9 scores were 0.738 and 0.571 for NC and SCDs groups, respectively. When we combined APOE ε4 and SCD-Q9 scores into the model, the AUC increased to 0.747 for discriminating OCI from NC. However, when OCI group was split into MCI and dementia groups, only total SCD-Q9 score was the independent affecting factor of MCI. Conclusion This study demonstrated that the distribution of APOE ε4 alleles did not vary with different demographic characteristics in a large-scale cohort from a memory clinic. APOE ε4 alleles may be associated with scores of SCD-Q9 reflecting the degree of cognitive complaints but their additional contribution to SCD-Q9 scores is marginal in discriminating between NC, SCDs, and OCI.

背景 主观认知下降(SCD)是阿尔茨海默病(AD)患者的早期表现之一,其早期简易筛查工具的研发有利于帮助临床早期识别AD,但部分轻度认知障碍(MCI)患者主观认知下降问卷9(SCD-Q9)为0分,可能与该问卷未涵盖所有SCD主诉有关.目的 对筛选出SCD-Q9的上位条目池即主观认知下降问卷21(SCD-Q21)进行汉化,找出可识别SCD风险的SCD主诉条目并进一步优化SCD-Q9.方法 2020-07-01至2020-08-31,参考国外问卷本土化标准程序,经原作者授权、同意后获得英文版原版问卷,采用Brislin"两人直译·回译"法对SCD-Q21进行汉化,并通过专家小组讨论及小样本预调查进行条目修订及文化调适.结果 本研究经严格的翻译、回译、对比回译、条目修订、文化调适及小样本预调查而形成中文版SCD-Q21终稿.中文版SCD-Q21终稿由21个条目组成,其中条目4、5、7、10、20、21为三分类选项,其余均为二分类选项;条目11、19为反向条目,其余均为正向条目;总分21分.

Chronic musculoskeletal pain is a common problem globally. Current evidence suggests that maladapted central pain pathways are associated with pain chronicity, for example, in postoperative pain after knee replacement. Other factors such as low mood, anxiety, and tendency to catastrophize are also important contributors. We aimed to investigate brain imaging features that underpin pain chronicity based on multivariate pattern analysis of cerebral blood flow (CBF), as a marker of maladaptive brain changes. This was achieved by identifying CBF patterns that discriminate chronic pain from pain-free conditions and by exploring their explanatory power for factors thought to drive pain chronification. In 44 chronic knee pain and 29 pain-free participants, we acquired both CBF and T1-weighted data. Participants completed questionnaires related to affective processes and pressure and cuff algometry to assess pain sensitization. Two factor scores were extracted from these scores representing negative affect and pain sensitization. A spatial covariance principal component analysis of CBF identified 5 components that significantly discriminated chronic pain participants from controls, with the unified network achieving 0.83 discriminatory accuracy (area under the curve). In chronic knee pain, significant patterns of relative hypoperfusion were evident in anterior default-mode and salience network hubs, while hyperperfusion was seen in posterior default mode, thalamus, and sensory regions. One component correlated positively with the pain sensitization score ( r = 0.43, P = 0.006), suggesting that this CBF pattern reflects neural activity changes encoding pain sensitization. Here, we report a distinct chronic knee pain-related representation of CBF, pointing toward a brain signature underpinning central aspects of pain sensitization.

Subjective cognitive decline (SCD) is regarded as the first clinical manifestation in the Alzheimer's disease (AD) continuum. Investigating populations with SCD is important for understanding the early pathological mechanisms of AD and identifying SCD-related biomarkers, which are critical for the early detection of AD. With the advent of advanced neuroimaging techniques, such as positron emission tomography (PET) and magnetic resonance imaging (MRI), accumulating evidence has revealed structural and functional brain alterations related to the symptoms of SCD. In this review, we summarize the main imaging features and key findings regarding SCD related to AD, from local and regional data to connectivity-based imaging measures, with the aim of delineating a multimodal imaging signature of SCD due to AD. Additionally, the interaction of SCD with other risk factors for dementia due to AD, such as age and the Apolipoprotein E (ApoE) ɛ4 status, has also been described. Finally, the possible explanations for the inconsistent and heterogeneous neuroimaging findings observed in individuals with SCD are discussed, along with future directions. Overall, the literature reveals a preferential vulnerability of AD signature regions in SCD in the context of AD, supporting the notion that individuals with SCD share a similar pattern of brain alterations with patients with mild cognitive impairment (MCI) and dementia due to AD. We conclude that these neuroimaging techniques, particularly multimodal neuroimaging techniques, have great potential for identifying the underlying pathological alterations associated with SCD. More longitudinal studies with larger sample sizes combined with more advanced imaging modeling approaches such as artificial intelligence are still warranted to establish their clinical utility.

Abstract Background Since SCD (plus) was standardized, little is known about its demographic characteristics and its outcomes of neuropsychological assessments, including the SCD questionnaire 9 (SCD‐Q9). Objective To characterize SCD (plus) by comparing the neuropsychological features among its subgroups and with normal controls (NC). Also, to explore its demographics and to understand the relation of the chief complaints and the scores of SCD‐Q9. Methods Multistage stratified cluster random sampling was conducted to select participants. As a result, 84 NC and 517 SCD (plus) were included. SCD (plus) was further classified into several subgroups (SCD‐C: concerned cognitive decline; SCD‐F: complaints about SCD within the past five years; SCD‐P: feeling performance being not as good as their peers; SCD+: presented> 3 of SCD (plus) features; SCD‐: presented ≤ 3 of SCD (plus) features (see the diagnostic criteria for the details)) and between‐group comparisons of neuropsychological scores were conducted. Point‐biserial correlation and binary logistic regression analyses were performed to investigate the demographic characteristics of its subgroups. Finally, Spearman correlation was used to better understand the relation of SCD (plus) to SCD‐Q9. Results (1) Scores of AVLT‐LR (AVLT‐LR: Auditory Verbal Learning Test‐Long Delayed Recall) and MoCA‐B (MoCA: Montreal Cognitive Assessment‐Basic) were lower in the SCD‐P group than those in the NC group, and the SCD+ group scored lower in the MoCA‐B and CDT(CDT: Clock Drawing Test) than the SCD‐ group. (2) Females were more concerned than male participants. Individuals with lower education level felt that their cognitive performance were worse than their peers. Also, younger people might express concerns more than the more elderly. People who had complaints of SCD‐P might be more likely to report SCD‐C, but less likely to report SCD‐F. (3) Positive correlations were found between the chief complaints of SCD (plus) and some items of SCD‐Q9. Conclusions SCD (plus) may be related to demographic factors. Individuals with SCD (plus) already exhibited cognitive impairment, which can be detected by SCD‐Q9.

AbstractChronic musculoskeletal pain is a common problem globally. Current evidence suggests that maladaptive modulation of central pain pathways is associated with pain chronicity following e.g. chronic post-operative pain after knee replacement. Other factors such as low mood, anxiety and tendency to catastrophize seem to also be important contributors. We aimed to identify a chronic pain brain signature that discriminates chronic pain from pain-free conditions using cerebral blood flow (CBF) measures, and explore how this signature relates to the chronic pain experience. In 44 chronic knee pain patients and 29 pain-free controls, we acquired CBF data (using arterial spin labelling) and T1-weighted images. Participants completed a series of questionnaires related to affective processes, and pressure and cuff algometry to assess pain sensitization. Two factor scores were extracted from these scores representing negative affect and pain sensitization, respectively. A spatial covariance principal components analysis of CBF identified five components that significantly discriminated chronic pain patients from controls, with the unified network achieving 0.83 discriminatory accuracy (area under the curve). In chronic knee pain, significant patterns of relative hypo-perfusion were evident in anterior regions of the default mode and salience network hubs, while hyperperfusion was seen in posterior default mode regions, the thalamus, and sensory regions. One component was positively correlated to the pain sensitization score (r=.43,p=.006), suggesting that this CBF pattern reflects the neural activity changes encoding pain sensitization. Here, we report the first chronic knee pain-related brain signature, pointing to a brain signature underpinning the central aspects of pain sensitisation.

Objective To investigate factors affecting the pattern of motor brain activation reported in people with Parkinson's (PwP), aiming to differentiate disease-specific features from treatment effects. Methods A co-ordinate-based-meta-analysis (CBMA) of functional motor neuroimaging studies involving patients with Parkinson's (PwP), and healthy controls (HC) identified 126 suitable articles. The experiments were grouped based on subject feature, medication status (onMed/offMed), deep brain stimulation (DBS) status (DBSon/DBSoff) and type of motor initiation. Results HC and PwP shared similar neural networks during upper extremity motor tasks but with differences of reported frequency in mainly bilateral putamen, insula and ipsilateral inferior parietal and precentral gyri. The activation height was significantly reduced in the bilateral putamen, left SMA, left subthalamus nucleus, right thalamus and right midial global pallidum in PwP(offMed)(vs. HC), and pre-SMA hypoactivation correlated with disease severity. These changes were not found in patients on dopamine replacement therapy (PwP(onMed)vs. HC) in line with a restorative function. By contrast, left SMA and primary motor cortex showed hyperactivation in the medicated state (vs. HC) suggesting dopaminergic overcompensation. Deep-brain stimulation (PwP during the high frequency subthalamus nucleus (STN) DBS vs. no stimulation) induced a decrease in left SMA activity and the expected increase in the left subthalamic/thalamic region regardless of hand movement. We further demonstrated a disease related effect of motor intention with only PwP(offMed)showing increased activation in the medial frontal lobe in self-initiated studies. Conclusion We describe a consistent disease-specific pattern of putaminal hypoactivation during motor tasks that appears reversed by dopamine replacement. Inconsistent reports of altered SMA/pre-SMA activation can be explained by task- and medication-specific variation in intention. Moreover, SMA activity was reduced during STN-DBS, while dopamine-induced hyperactivation of SMA which might underpin hyperdynamic L-dopa related overcompensation.

BACKGROUND:Since subjective cognitive decline (SCD) was standardized in 2014, many studies have investigated its features. However, the risk of SCD (plus) progressing to AD is much higher, and yet there have been few studies reporting the risk factors and neuropsychological assessment characteristics of SCD (plus).OBJECTIVE:To characterize SCD (plus) by comparing it with normal control (NC), amnesic mild cognitive impairment (aMCI), and Alzheimer Disease (AD) regarding their demographics, lifestyle, family history of dementia, multimorbidity and the neuropsychological assessments.METHODS:A total of 135 participants were recruited, including 23 NC, 30 SCD (plus), 45 aMCI and 37 AD. Descriptive statistics were provided. A logistic regression model was used to analyze the affecting factors of SCD (plus), and finally the Receiver Operating Characteristic (ROC) analysis was applied to distinguish between SCD (plus) and NC.RESULTS:(1) SCD (plus) group was younger than both the aMCI group and AD group. It consisted of more participants with mental work and higher body mass index (BMI) than the AD group. (2) Scores of Auditory Verbal Learning Test - Immediate recall (AVLT-IR) and AVLT-Long delayed recall (AVLT-LR) decreased in the following order: NC→SCD (plus)→aMCI→AD. (3) The Area Under Curve (AUC) for discriminating SCD (plus) and NC group was from 0.673 to 0.838.CONCLUSION:Aging is an important risk factor of both NC progressing to SCD (plus), and SCD (plus) progressing to aMCI or AD. In addition to aging, lower education level and lower BMI were significantly associated with greater odds of SCD (plus) progressing to aMCI or AD patients, whereas mental work was a protective factor of SCD (plus) progressing to AD. Finally, AVLT is a sensitive indicator of the cognitive decline and impairment in SCD (plus) in relative to normal controls.

In this paper, we present a generic deep convolutional neural network (DCNN) for multi-class image segmentation. It is based on a well-established supervised end-to-end DCNN model, known as U-net. U-net is firstly modified by adding widely used batch normalization and residual block (named as BRU-net) to improve the efficiency of model training. Based on BRU-net, we further introduce a dynamically weighted cross-entropy loss function. The weighting scheme is calculated based on the pixel-wise prediction accuracy during the training process. Assigning higher weights to pixels with lower segmentation accuracies enables the network to learn more from poorly predicted image regions. Our method is named as feedback weighted U-net (FU-net). We have evaluated our method based on T1- weighted brain MRI for the segmentation of midbrain and substantia nigra, where the number of pixels in each class is extremely unbalanced to each other. Based on the dice coefficient measurement, our proposed FU-net has outperformed BRU-net and U-net with statistical significance, especially when only a small number of training examples are available. The code is publicly available in GitHub (GitHub link: https://github.com/MinaJf/FU-net).