Study of the Influence of Age in 18F-FDG PET Images Using a Data-Driven Approach and Its Evaluation in Alzheimer's Disease

Evaluating the effect of rapamycin treatment in Alzheimer's disease and aging using in vivo imaging: the ERAP phase IIa clinical study protocol

BACKGROUND

Rapamycin is an inhibitor of the mechanistic target of rapamycin (mTOR) protein kinase, and preclinical data demonstrate that it is a promising candidate for a general gero- and neuroprotective treatment in humans. Results from mouse models of Alzheimer's disease have shown beneficial effects of rapamycin, including preventing or reversing cognitive deficits, reducing amyloid oligomers and tauopathies and normalizing synaptic plasticity and cerebral glucose uptake. The "Evaluating Rapamycin Treatment in Alzheimer's Disease using Positron Emission Tomography" (ERAP) trial aims to test if these results translate to humans through evaluating the change in cerebral glucose uptake following six months of rapamycin treatment in participants with early-stage Alzheimer's disease.

METHODS

ERAP is a six-month-long, single-arm, open-label, phase IIa biomarker-driven study evaluating if the drug rapamycin can be repurposed to treat Alzheimer's disease. Fifteen patients will be included and treated with a weekly dose of 7 mg rapamycin for six months. The primary endpoint will be change in cerebral glucose uptake, measured using [18F]FDG positron emission tomography. Secondary endpoints include changes in cognitive measures, markers in cerebrospinal fluid as well as cerebral blood flow measured using magnetic resonance imaging. As exploratory outcomes, the study will assess change in multiple age-related pathological processes, such as periodontal inflammation, retinal degeneration, bone mineral density loss, atherosclerosis and decreased cardiac function.

DISCUSSION

The ERAP study is a clinical trial using in vivo imaging biomarkers to assess the repurposing of rapamycin for the treatment of Alzheimer's disease. If successful, the study would provide a strong rationale for large-scale evaluation of mTOR-inhibitors as a potential disease-modifying treatment in Alzheimer's disease.

TRIAL REGISTRATION

ClinicalTrials.gov ID NCT06022068, date of registration 2023-08-30.

Distance-weighted Sinkhorn loss for Alzheimer's disease classification

Traditional loss functions such as cross-entropy loss often quantify the penalty for each mis-classified training sample without adequately considering its distance from the ground truth class distribution in the feature space. Intuitively, the larger this distance is, the higher the penalty should be. With this observation, we propose a penalty called distance-weighted Sinkhorn (DWS) loss. For each mis-classified training sample (with predicted label A and true label B), its contribution to the DWS loss positively correlates to the distance the training sample needs to travel to reach the ground truth distribution of all the A samples. We apply the DWS framework with a neural network to classify different stages of Alzheimer's disease. Our empirical results demonstrate that the DWS framework outperforms the traditional neural network loss functions and is comparable or better to traditional machine learning methods, highlighting its potential in biomedical informatics and data science.

The remodeling of metabolic brain pattern in patients with extracranial diffuse large B-cell lymphoma

BACKGROUND

Owing to the advances in diagnosis and therapy, survival or remission rates for lymphoma have improved prominently. Apart from the lymphoma- and chemotherapy-related somatic symptom burden, increasing attention has been drawn to the health-related quality of life. The application of 18F-fluorodeoxyglucose positron emission tomography-computed tomography (18F-FDG PET/CT) has been routinely recommended for the staging and response assessment of FDG-avid lymphoma. However, up till now, only a few researches have investigated the brain metabolic impairments in patients with pre-treatment lymphoma. The determination of the lymphoma-related metabolic brain pattern would facilitate exploring the tailored therapeutic regimen to alleviate not only the physiological, but also the psychological symptoms. In this retrospective study, we aimed to establish the diffuse large B-cell lymphoma-related pattern (DLBCLRP) of metabolic brain network and investigate the correlations between DLBCLRP and several indexes of the staging and response assessment.

RESULTS

The established DLBCLRP was characterized by the increased metabolic activity in bilateral cerebellum, brainstem, thalamus, striatum, hippocampus, amygdala, parahippocampal gyrus and right middle temporal gyrus and by the decreased metabolic activity in bilateral occipital lobe, parietal lobe, anterior cingulate gyrus, midcingulate cortex and medial frontal gyrus. Significant difference in the baseline expression of DLBCLRP was found among complete metabolic response (CMR), partial metabolic response (PMR) and progressive metabolic disease (PMD) groups (P < 0.01). DLBCLRP expressions were also significantly or tended to be positively correlated with international prognostic index (IPI) (rs = 0.306, P < 0.05), lg(total metabolic tumor volume, TMTV) (r = 0.298, P < 0.05) and lg(total lesion glycolysis, TLG) (r = 0.233, P = 0.064). Though no significant correlation of DLBCLRP expression was found with Ann Arbor staging or tumor SUVmax (P > 0.05), the post-treatment declines of DLBCLRP expression were significantly positively correlated with Ann Arbor staging (rs = 0.284, P < 0.05) and IPI (rs = 0.297, P < 0.05).

CONCLUSIONS

The proposed DLBCLRP would lay the foundation for further investigating the cerebral dysfunction related to DLBCL itself and/or treatments. Besides, the expression of DLBCLRP was associated with the tumor burden of lymphoma, implying a potential biomarker for prognosis.

Multi-Group Tensor Canonical Correlation Analysis

Tensor Canonical Correlation Analysis (TCCA) is a commonly employed statistical method utilized to examine linear associations between two sets of tensor datasets. However, the existing TCCA models fail to adequately address the heterogeneity present in real-world tensor data, such as brain imaging data collected from diverse groups characterized by factors like sex and race. Consequently, these models may yield biased outcomes. In order to surmount this constraint, we propose a novel approach called Multi-Group TCCA (MG-TCCA), which enables the joint analysis of multiple subgroups. By incorporating a dual sparsity structure and a block coordinate ascent algorithm, our MG-TCCA method effectively addresses heterogeneity and leverages information across different groups to identify consistent signals. This novel approach facilitates the quantification of shared and individual structures, reduces data dimensionality, and enables visual exploration. To empirically validate our approach, we conduct a study focused on investigating correlations between two brain positron emission tomography (PET) modalities (AV-45 and FDG) within an Alzheimer's disease (AD) cohort. Our results demonstrate that MG-TCCA surpasses traditional TCCA in identifying sex-specific cross-modality imaging correlations. This heightened performance of MG-TCCA provides valuable insights for the characterization of multimodal imaging biomarkers in AD.

Adjustment for the Age- and Gender-Related Metabolic Changes Improves the Differential Diagnosis of Parkinsonism

Age and gender are the important factors for brain metabolic declines in both normal aging and neurodegeneration, and the confounding effects may influence early and differential diagnosis of neurodegenerative diseases based on the [18F]fluorodeoxyglucose positron emission tomography ([18F]FDG PET). We aimed to explore the potential of the adjustment of age- and gender-related confounding factors on [18F]FDG PET images in differentiation of Parkinson's disease (PD), multiple system atrophy (MSA) and progressive supra-nuclear palsy (PSP). Eight hundred and seventy-seven clinically definitely diagnosed Parkinsonian patients from a benchmark Huashan Parkinsonian PET imaging database were included. An age- and gender-adjusted Z (AGAZ) score was established based on the gender-specific longitudinal metabolic changes on healthy subjects. AGAZ scores and standardized uptake value ratio (SUVR) values were quantified at regional-level and support vector machine-based error-correcting output codes method was applied for classification. Additional references of the classifications based on metabolic pattern scores were included. The feature-based AGAZ score showed the best performance in classification (accuracy for PD, MSA, PSP: 93.1%, 96.3%, 94.8%). In both genders, the AGAZ score consistently achieved the best efficiency, and the improvements compared to the conventional SUVR value for PD, MSA, and PSP mainly laid in specificity (Male: 5.7%; Female: 11.1%), sensitivity (Male: 7.2%; Female: 7.3%), and sensitivity (Male: 7.3%; Female: 17.2%). Female patients benefited more from the adjustment on [18F]FDG PET in MSA and PSP groups (absolute net reclassification index, p < 0.001). Collectively, the adjustment of age- and gender-related confounding factors may improve the differential diagnosis of Parkinsonism. Particularly, the diagnosis of female Parkinsonian population has the best improvement from this correction.

Supplementary Information

The online version contains supplementary material available at 10.1007/s43657-022-00079-6.

Associating brain imaging phenotypes and genetic risk factors via a hypergraph based netNMF method

Abstract

Alzheimer's disease (AD) is a severe neurodegenerative disease for which there is currently no effective treatment. Mild cognitive impairment (MCI) is an early disease that may progress to AD. The effective diagnosis of AD and MCI in the early stage has important clinical significance.

Methods

To this end, this paper proposed a hypergraph-based netNMF (HG-netNMF) algorithm for integrating structural magnetic resonance imaging (sMRI) of AD and MCI with corresponding gene expression profiles.

Results

Hypergraph regularization assumes that regions of interest (ROIs) and genes were located on a non-linear low-dimensional manifold and can capture the inherent prevalence of two modalities of data and mined high-order correlation features of the two data. Further, this paper used the HG-netNMF algorithm to construct a brain structure connection network and a protein interaction network (PPI) with potential role relationships, mine the risk (ROI) and key genes of both, and conduct a series of bioinformatics analyses.

Conclusion

Finally, this paper used the risk ROI and key genes of the AD and MCI groups to construct diagnostic models. The AUC of the AD group and MCI group were 0.8 and 0.797, respectively.

Effects of donepezil on the amplitude of low-frequency fluctuations in the brain of patients with Alzheimer's disease: evidence from resting-state functional magnetic resonance imaging

OBJECTIVES

To monitor the effects of donepezil on spontaneous neuronal activity (SNA), and the mechanisms underlying these effects in patients with mild-to-moderate Alzheimer's disease, using the amplitude of low-frequency fluctuations (ALFFs), a metric of resting-state functional MRI (rs-fMRI).

METHODS

Eleven patients with Alzheimer's disease were treated with donepezil for 6 months. Before and after treatment, the Mini-Mental State Examination (MMSE), Alzheimer's Disease Assessment Scale-Cognitive Subscale (ADAS-cog), Clinical Dementia Rating (CDR), Neuropsychiatric Inventory and Activities of Daily Living scores, along with rs-fMRI of patients were assessed. Eleven age-, sex-, and education-matched controls underwent MMSE and CDR assessments and rs-fMRI at enrollment. The ALFFs of the whole brain were obtained and compared between the groups.

RESULTS

Following donepezil treatment, MMSE scores increased (P = 0.043) and ADAS-cog scores decreased (P = 0.010). Regarding SNA post-treatment, ALFF increased significantly in the right triangular part of the inferior frontal gyrus (IFGtriang; P = 0.030; d = -0.595) and the right orbital part of the inferior frontal gyrus (P = 0.044; d = -0.628) and decreased significantly in the left medial orbital part of the superior frontal gyrus (P = 0.039; d = 0.606) and the right gyrus rectus (P = 0.010; d = 0.609). Furthermore, the changes in ADAS-cog scores from before to after treatment were positively correlated with the changes in ALFF in the right IFGtriang (r = 0.645; P = 0.032).

CONCLUSIONS

Donepezil improved SNA in the frontal lobe of patients with Alzheimer's disease. Therefore, ALFF was demonstrated to be a potential tool for assessing the effectiveness of Alzheimer's disease treatment.

Implication of metabolic and dopamine transporter PET in dementia with Lewy bodies

To evaluate the implication of ¹⁸F-fluorodeoxyglucose (FDG)- and dopamine transporter (DAT)-positron emission tomography (PET) in the diagnosis and clinical symptoms of dementia with Lewy bodies (DLB), 55 DLB patients and 49 controls underwent neuropsychological evaluation and FDG-, DAT-, and ¹⁸F-Florbetaben (FBB) PET. DAT- and FDG-uptake and FDG/DAT ratio were measured in the anterior and posterior striatum. The first principal component (PC1) of FDG subject residual profiles was identified for each subject. Receiver operating characteristic curve analyses for the diagnosis of DLB were performed using FDG- and DAT-PET biomarkers as predictors, and general linear models for motor severity and cognitive scores were performed adding FBB standardized uptake value ratio as a predictor. Increased metabolism in the bilateral putamen, vermis, and somato-motor cortices, which characterized PC1, was observed in the DLB group, compared to the control group. A combination of posterior putamen FDG/DAT ratio and PC1 showed the highest diagnostic accuracy (91.8% sensitivity and 96.4% specificity), which was significantly greater than that obtained by DAT uptake alone. Striatal DAT uptake and PC1 independently contributed to motor severity and language, memory, frontal/executive, and general cognitive dysfunction in DLB patients, while only PC1 contributed to attention and visuospatial dysfunction.

Imaging Techniques in Alzheimer's Disease: A Review of Applications in Early Diagnosis and Longitudinal Monitoring

BACKGROUND

Alzheimer's disease (AD) is a progressive neurodegenerative disorder affecting many individuals worldwide with no effective treatment to date. AD is characterized by the formation of senile plaques and neurofibrillary tangles, followed by neurodegeneration, which leads to cognitive decline and eventually death.

INTRODUCTION

In AD, pathological changes occur many years before disease onset. Since disease-modifying therapies may be the most beneficial in the early stages of AD, biomarkers for the early diagnosis and longitudinal monitoring of disease progression are essential. Multiple imaging techniques with associated biomarkers are used to identify and monitor AD.

AIM

In this review, we discuss the contemporary early diagnosis and longitudinal monitoring of AD with imaging techniques regarding their diagnostic utility, benefits and limitations. Additionally, novel techniques, applications and biomarkers for AD research are assessed.

FINDINGS

Reduced hippocampal volume is a biomarker for neurodegeneration, but atrophy is not an AD-specific measure. Hypometabolism in temporoparietal regions is seen as a biomarker for AD. However, glucose uptake reflects astrocyte function rather than neuronal function. Amyloid-β (Aβ) is the earliest hallmark of AD and can be measured with positron emission tomography (PET), but Aβ accumulation stagnates as disease progresses. Therefore, Aβ may not be a suitable biomarker for monitoring disease progression. The measurement of tau accumulation with PET radiotracers exhibited promising results in both early diagnosis and longitudinal monitoring, but large-scale validation of these radiotracers is required. The implementation of new processing techniques, applications of other imaging techniques and novel biomarkers can contribute to understanding AD and finding a cure.

CONCLUSIONS

Several biomarkers are proposed for the early diagnosis and longitudinal monitoring of AD with imaging techniques, but all these biomarkers have their limitations regarding specificity, reliability and sensitivity. Future perspectives. Future research should focus on expanding the employment of imaging techniques and identifying novel biomarkers that reflect AD pathology in the earliest stages.

Disappearing metabolic youthfulness in the cognitively impaired female brain

Sex differences play a vital role in human brain structure and physiology. Previous reports have proposed evidence hinting at a metabolic advantage in female brains across adulthood. It remained to be determined whether this advantage would be maintained across the spectrum of cognitive impairment, up to and including dementia due to Alzheimer's disease (AD). Here, using a machine-learning algorithm, we explore sex differences in metabolic brain-age derived from fluorodeoxyglucose positron emission tomography imaging among cognitively healthy individuals and those affected by mild cognitive impairment and clinically probable AD. First, we report that cognitively healthy male participants showed a persistently "older" looking brains when compared to healthy female participants in term of metabolic brain age, confirming earlier reports. However, this distinction disappeared among MCI individuals and probable AD patients, and this loss could not be explained by an accompanying neurodegeneration. This would seem to indicate that females have a higher rate of decline in brain glucose metabolism when cognitively impaired to negate their prior advantage.

FDG-PET Profiles of Extratemporal Metabolism as a Predictor of Surgical Failure in Temporal Lobe Epilepsy

Objective: Metabolic abnormality in the extratemporal area on fluorine-18-fluorodeoxyglucose positron emission tomography (FDG-PET) is not an uncommon finding in drug-resistant temporal lobe epilepsy (TLE), however the correlation between extratemporal metabolic abnormalities and surgical long-term prognosis has not been fully elucidated. We aim to investigate FDG-PET extratemporal metabolic profiles predictive of failure in surgery for TLE patients. Methods: Eighty-two patients with unilateral TLE (48 female, 34 male; 25.6 ± 10.6 years old; 37 left TLE, 45 right TLE) and 30 healthy age-matched controls were enrolled. Patients were classified either as experiencing seizure-recurrence (SZR, Engel class II through IV) or seizure-free (SZF, Engel class I) at least 1 year after surgery. Regional cerebral metabolism was evaluated by FDG-PET with statistical parametric mapping (SPM12). Abnormal metabolic profiles and patterns on FDG-PET in SZR group were evaluated and compared with those of healthy control and SZF subjects on SPM12. Volume and intensity as well as special brain areas of abnormal metabolism in temporal and extratemporal regions were quantified and visualized. Results: With a median follow-up of 1.5 years, 60% of patients achieved Engel class I (SZF). SZR was associated with left TLE and widespread hypometabolism in FDG-PET visual assessment (both p < 0.05). All patients had hypometabolism in the ipsilateral temporal lobe but SZR was not correlated with volume or intensity of temporal hypometabolism (median, 1,456 vs. 1,040 mm3; p > 0.05). SZR was correlated with extratemporal metabolic abnormalities that differed according to lateralization: in right TLE, SZR exhibited larger volume in extratemporal areas compared to SZF (median, 11,060 vs. 2,112 mm3; p < 0.05). Surgical failure was characterized by Cingulum_Ant_R/L, Frontal_Inf_Orb_R abnormal metabolism in extratemporal regions. In left TLE, SZR presented a larger involvement of extratemporal areas similar to right TLE but with no significant (median, 5,873 vs. 3,464 mm3; p > 0.05), Cingulum_Ant_ R/L, Parietal_Inf_L, Postcentral_L, and Precuneus_R involved metabolic abnormalities were correlated with SZR. Conclusions: Extratemporal metabolic profiles detected by FDG-PET may indicate a prominent cause of TLE surgery failure and should be considered in predictive models for epilepsy surgery. Seizure control after surgery might be improved by investigating extratemporal areas as candidates for resection or neuromodulation.

F-18 FDG PET brain imaging in symptomatic arthroprosthetic cobaltism

PURPOSE

Imaging studies of cobalt toxicity from cobalt-chromium alloy arthroprosthetics have focused on the local intra-articular and peri-articular presentation from failing joint replacements. Most studies investigating neurological findings have been small case series focused on the clinical findings of memory loss, diminished executive function, tremor, hearing and vision loss, depression, and emotional lability. This study utilizes software-based quantitative analysis of brain metabolism to assess the degree of hypometabolism and areas of susceptibility, determine if a pattern of involvement exists, and measure reversibility of findings after prosthetic revision to cobalt-free appliances.

METHODS

Over 48 months, 247 consecutive patients presenting to an orthopedic clinic with an arthroprosthetic joint containing any cobalt-chromium part were screened with whole blood and urine cobalt levels. A clinically validated inventory of 10 symptoms was obtained. Symptomatic patients with a blood cobalt level above 0.4 mcg/L or urine cobalt greater than 1 mcg/L underwent F-18 FDG PET brain imaging. Analysis was performed with FDA-approved quantitative brain analysis software with the pons as the reference region. Control group was the normal brain atlas within the software.

RESULTS

Of the 247 consecutively screened patients, 123 had blood and urine cobalt levels above the threshold. The 69 scanned patients had statistically significant regional hypometabolism and higher symptoms inventory. Fifty-seven patients were retained in the study. Distribution of hypometabolism was in descending order: temporal, frontal, Broca's areas, anterior cingulate, parietal, posterior cingulate, visual, sensorimotor, thalamic, and lastly caudate. Metal-on-metal (MoM) and metal-on-plastic (MoP) joint replacements produced similar patterns of hypometabolism. Of 15 patients with necessary revision surgery, 8 demonstrated improved metabolism when later re-scanned.

CONCLUSION

All scanned patients had regions of significant hypometabolism. Neurological toxicity from elevated systemic cobalt levels following arthroprosthetic joint replacement has a pattern of regional susceptibility similar to heavy metals and solvents, differing from classical dementias and may occur at blood and urine cobalt levels as low as 0.4 mcg/L and 1 mcg/L, respectively. Presently accepted thresholds for cobalt exposure and monitoring may need revision. Quantitative F-18 FDG PET brain imaging may aid in the decision process for treatment options and timing of possible medical versus surgical intervention.

Selection of the optimal intensity normalization region for FDG-PET studies of normal aging and Alzheimer's disease

The primary method for measuring brain metabolism in humans is positron emission tomography (PET) imaging using the tracer 18F-fluorodeoxyglucose (FDG). Standardized uptake value ratios (SUVR) are commonly calculated from FDG-PET images to examine intra- and inter-subject effects. Various reference regions are used in the literature of FDG-PET studies of normal aging, making comparison between studies difficult. Our primary objective was to determine the optimal SUVR reference region in the context of healthy aging, using partial volume effect (PVE) and non-PVE corrected data. We calculated quantitative cerebral metabolic rates of glucose (CMRg) from PVE-corrected and non-corrected images from young and older adults. We also investigated regional atrophy using magnetic resonance (MR) images. FreeSurfer 6.0 atlases were used to explore possible reference regions of interest (ROI). Multiple regression was used to predict CMRg data, in each FreeSurfer ROI, with age and sex as predictors. Age had the least effect in predicting CMRg for PVE corrected data in the pons (r2 = 2.83 × 10-3, p = 0.67). For non-PVE corrected data age also had the least effect in predicting CMRg in the pons (r2 = 3.12 × 10-3, p = 0.67). We compared the effects of using the whole brain or the pons as a reference region in PVE corrected data in two regions susceptible to hypometabolism in Alzheimer's disease, the posterior cingulate and precuneus. Using the whole brain as a reference region resulted in non-significant group differences in the posterior cingulate while there were significant differences between all three groups in the precuneus (all p < 0.004). When using the pons as a reference region there was significant differences between all groups for both the posterior cingulate and the precuneus (all p < 0.001). Therefore, the use of the pons as a reference region is more sensitive to hypometabism changes associated with Alzheimer's disease than the whole brain.

The risk prediction of Alzheimer's disease based on the deep learning model of brain 18F-FDG positron emission tomography

In order to predict the risks of Alzheimer's Disease (AD) based on the deep learning model of brain 18F-FDG positron emission tomography (PET), a total of 350 mild cognitive impairment (MCI) participants from the Alzheimer's Disease Neuroimaging Initiative (ADNI) database were selected as the research objects; in addition, the Convolutional Architecture for Fast Feature Embedding (CAFFE) was selected as the framework of the deep learning platform; the FDG PET image features of each participant were extracted by a deep convolution network model to construct the prediction and classification models; therefore, the MCI stage features were classified and the transformation was predicted. The results showed that in terms of the MCI transformation prediction, the sensitivity and specificity of conv3 classification were respectively 91.02% and 77.63%; in terms of the Late Mild Cognitive Impairment (LMCI) and Early Mild Cognitive Impairment (EMCI) classification, the accuracy of conv5 classification was 72.19%, and the sensitivity and specificity of conv5 were all 73% approximately. Thus, it was seen that the model constructed in the research could be used to solve the problems of MCI transformation prediction, which also had certain effects on the classifications of EMCI and LMCI. The risk prediction of AD based on the deep learning model of brain 18F-FDG PET discussed in the research matched the expected results. It provided a relatively accurate reference model for the prediction of AD. Despite the deficiencies of the research process, the research results have provided certain references and guidance for the future exploration of accurate AD prediction model; therefore, the research is of great significance.

Berkson's paradox and weighted distributions: An application to Alzheimer's disease

One reason for observing in practice a false positive or negative correlation between two random variables, which are either not correlated or correlated with a different direction, is the overrepresentation in the sample of individuals satisfying specific properties. In 1946, Berkson first illustrated the presence of a false correlation due to this last reason, which is known as Berkson's paradox and is one of the most famous paradox in probability and statistics. In this paper, the concept of weighted distributions is utilized to describe Berskon's paradox. Moreover, a proper procedure is suggested to make inference for the population given a biased sample which possesses all the characteristics of Berkson's paradox. A real data application for patients with dementia due to Alzheimer's disease demonstrates that the proposed method reveals characteristics of the population that are masked by the sampling procedure.

Bias-adjustment in neuroimaging-based brain age frameworks: A robust scheme

The level of prediction error in the brain age estimation frameworks is associated with the authenticity of statistical inference on the basis of regression models. In this paper, we present an efficacious and plain bias-adjustment scheme using chronological age as a covariate through the training set for downgrading the prediction bias in a Brain-age estimation framework. We applied proposed bias-adjustment scheme coupled by a machine learning-based brain age framework on a large set of metabolic brain features acquired from 675 cognitively unimpaired adults through fluorodeoxyglucose positron emission tomography data as the training set to build a robust Brain-age estimation framework. Then, we tested the reliability of proposed bias-adjustment scheme on 75 cognitively unimpaired adults, 561 mild cognitive impairment patients as well as 362 Alzheimer's disease patients as independent test sets. Using the proposed method, we gained a strong R2 of 0.81 between the chronological age and brain estimated age, as well as an excellent mean absolute error of 2.66 years on 75 cognitively unimpaired adults as an independent set; whereas an R2 of 0.24 and a mean absolute error of 4.71 years was achieved without bias-adjustment. The simulation results demonstrated that the proposed bias-adjustment scheme has a strong capability to diminish prediction error in brain age estimation frameworks for clinical settings.