A Machine Learning-Based Holistic Approach to Predict the Clinical Course of Patients within the Alzheimer’s Disease Spectrum

Background: Alzheimer’s disease (AD) is a neurodegenerative condition driven by multifactorial etiology. Mild cognitive impairment (MCI) is a transitional condition between healthy aging and dementia. No reliable biomarkers are available to predict the conversion from MCI to AD. Objective: To evaluate the use of machine learning (ML) on a wealth of data offered by the Alzheimer’s Disease Neuroimaging Initiative (ADNI) and Alzheimer’s Disease Metabolomics Consortium (ADMC) database in the prediction of the MCI to AD conversion. Methods: We implemented an ML-based Random Forest (RF) algorithm to predict conversion from MCI to AD. Data related to the study population (587 MCI subjects) were analyzed by RF as separate or combined features and assessed for classification power. Four classes of variables were considered: neuropsychological test scores, AD-related cerebrospinal fluid (CSF) biomarkers, peripheral biomarkers, and structural magnetic resonance imaging (MRI) variables. Results: The ML-based algorithm exhibited 86% accuracy in predicting the AD conversion of MCI subjects. When assessing the features that helped the most, neuropsychological test scores, MRI data, and CSF biomarkers were the most relevant in the MCI to AD prediction. Peripheral parameters were effective when employed in association with neuropsychological test scores. Age and sex differences modulated the prediction accuracy. AD conversion was more effectively predicted in females and younger subjects. Conclusion: Our findings support the notion that AD-related neurodegenerative processes result from the concerted activity of multiple pathological mechanisms and factors that act inside and outside the brain and are dynamically affected by age and sex.

Download Full-text

A machine learning-based holistic approach for diagnoses within the Alzheimer's disease spectrum

10.1101/2020.10.02.20205559 ◽

2020 ◽

Author(s):

Noemi Massetti ◽

Alberto Granzotto ◽

Manuela Bomba ◽

Stefano Delli Pizzi ◽

Alessandra Mosca ◽

...

Keyword(s):

Machine Learning ◽

Alzheimer’S Disease ◽

Alzheimer's Disease ◽

Healthy Aging ◽

Holistic Approach ◽

Structural Mri ◽

Relative Contribution ◽

Disease Spectrum ◽

Organ Specific ◽

The Central Nervous System

Alzheimer's disease (AD) is a neurodegenerative condition driven by a multifactorial etiology. We employed a machine learning (ML) based algorithm and the wealth of information offered by the Alzheimer's Disease Neuroimaging Initiative (ADNI) database to investigate the relative contribution of clinically relevant factors for identifying subjects affected by Mild Cognitive Impairment (MCI), a transitional status between healthy aging and dementia. Our ML-based Random Forest (RF) algorithm did not help predict clinical outcomes and the AD conversion of MCI subjects. On the other hand, non-converting (ncMCI) subjects were correctly classified and predicted. Two neuropsychological tests, the FAQ and ADAS13, were the most relevant features used for the classification and prediction of younger, under 70, ncMCI subjects. Structural MRI data combined with systemic parameters and the cardiovascular status were instead the most critical factors for the classification of over 70 ncMCI subjects. Our results support the notion that AD is not an organ-specific condition and results from pathological processes inside and outside the Central Nervous System.

Download Full-text

Using Machine Learning to Differentiate between Healthy Aging, Mild Cognitive Impairment, & Alzheimer's Disease

10.17918/d8th5x ◽

2021 ◽

Author(s):

Monica Truelove-Hill

Keyword(s):

Machine Learning ◽

Alzheimer’S Disease ◽

Alzheimer's Disease ◽

Cognitive Impairment ◽

Mild Cognitive Impairment ◽

Healthy Aging

Download Full-text

Value of Neuropsychological Tests to Identify Patients with Depressive Symptoms on the Alzheimer’s Disease Continuum

Journal of Alzheimer s Disease ◽

10.3233/jad-200710 ◽

2020 ◽

Vol 78 (2) ◽

pp. 819-826

Author(s):

Felix Menne ◽

Carola Gertrud Schipke ◽

Arne Klostermann ◽

Manuel Fuentes-Casañ ◽

Silka Dawn Freiesleben ◽

...

Keyword(s):

Alzheimer’S Disease ◽

Alzheimer's Disease ◽

Depressive Symptoms ◽

Neuropsychological Tests ◽

Neuropsychological Test ◽

Memory Task ◽

Geriatric Depression Scale ◽

Depression Scale ◽

Area Under The Curve ◽

Csf Biomarkers

Background: Depressive symptoms often co-occur with Alzheimer’s disease (AD) and can impact neuropsychological test results. In early stages of AD, disentangling cognitive impairments due to depression from those due to neurodegeneration often poses a challenge. Objective: We aimed to identify neuropsychological tests able to detect AD-typical pathology while taking into account varying degrees of depressive symptoms. Methods: A battery of neuropsychological tests (CERAD-NP) and the Geriatric Depression Scale (GDS) were assessed, and cerebrospinal fluid (CSF) biomarkers were obtained. After stratifying patients into CSF positive or negative and into low, moderate, or high GDS score groups, sensitivity and specificity and area under the curve (AUC) were calculated for each subtest. Results: 497 participants were included in the analyses. In patients with low GDS scores (≤10), the highest AUC (0.72) was achieved by Mini-Mental State Examination, followed by Constructional Praxis Recall and Wordlist Total Recall (AUC = 0.714, both). In patients with moderate (11–20) and high (≥21) GDS scores, Trail Making Test-B (TMT-B) revealed the highest AUCs with 0.77 and 0.82, respectively. Conclusion: Neuropsychological tests showing AD-typical pathology in participants with low GDS scores are in-line with previous results. In patients with higher GDS scores, TMT-B showed the best discrimination. This indicates the need to focus on executive function rather than on memory task results in depressed patients to explore a risk for AD.

Download Full-text

0057 Actigraphy-Based Circadian Measures and Cerebrospinal Fluid Biomarkers of Neurodegeneration in Alzheimer’s Disease with Mild Cognitive Impairment

SLEEP ◽

10.1093/sleep/zsaa056.055 ◽

2020 ◽

Vol 43 (Supplement_1) ◽

pp. A23-A23

Author(s):

R Mehra ◽

R Bhambra ◽

J Bena ◽

L Bekris ◽

J Leverenz ◽

...

Keyword(s):

Alzheimer’S Disease ◽

Alzheimer's Disease ◽

Cerebrospinal Fluid ◽

Cognitive Impairment ◽

Mild Cognitive Impairment ◽

Csf Biomarkers ◽

Significance Level ◽

Unit Increase ◽

T Tau ◽

Age And Sex

Abstract Introduction Although recent data implicates sleep and circadian disruption to neurodegeneration in Alzheimer’s Disease (AD), the association of objective circadian biomarkers and neurodegeneration remains understudied. We hypothesize that actigraphy-based circadian measures are associated with cerebrospinal fluid (CSF) biomarkers of neurodegeneration in those mild cognitive impairment due to AD (MCI-AD). Methods Eighteen patients with CSF biomarker-confirmed MCI-AD underwent actigraphy monitoring generating the following circadian measures: amplitude, F-ratio and mesor and morning collection of CSF biomarkers of neurodegeneration (Aβ42,t-tau,p-tau). Linear models were used to evaluate the association of circadian and CSF measures; logarithmic transformations were performed on neurodegenerative markers for greater normality. Analysis was performed using SAS software. A significance level of 0.05 was assumed for all tests. Results Eighteen MCI-AD patients who were 68± 6.2 years, 44% female, with median AHI=12 and underwent actigraphy monitoring for 8.2+/-3.2 days were included. There was no significant association of circadian measures and Aβ42 nor with mesor and neurodegeneration biomarkers. Amplitude was associated with both p-tau and t-tau, such that each 10 unit increase in amplitude resulted in a predicted increase in p-tau of 8% (95% CI:1%-15%, p=0.018) and an increase of 13% (3%-23%; p=0.01) in t-tau. F-ratio was positively associated with p-tau and t-tau; each 1000 unit increase in F-ratio resulted in a predicted 12% (4%-22%; p=0.007) increase in P-tau and 20%(6%-35%; p=0.005) increase in t-tau. Associations of these circadian measures and CSF levels of p-tau and t-tau remained statistically significant after adjustment for age and sex. Conclusion Among patients with symptomatic MCI stages of AD, objective measures of circadian rhythm disruption are associated with CSF-based biomarkers of neurodegeneration even after consideration of age and sex. Future investigation should clarify directionality of this association and potential utility of circadian-based interventions in the mitigation of AD progression. Support N/A

Download Full-text

A metabolite-based machine learning approach to diagnose Alzheimer’s-type dementia in blood: Results from the European Medical Information Framework for Alzheimer’s Disease biomarker discovery cohort

10.1101/19007146 ◽

2019 ◽

Author(s):

Daniel Stamate ◽

Min Kim ◽

Petroula Proitsi ◽

Sarah Westwood ◽

Alison Baird ◽

...

Keyword(s):

Machine Learning ◽

Alzheimer’S Disease ◽

Alzheimer's Disease ◽

Medical Information ◽

Biomarker Discovery ◽

Gradient Boosting ◽

Plasma Metabolites ◽

Csf Biomarkers ◽

Extreme Gradient Boosting ◽

And Gender

AbstractINTRODUCTIONMachine learning (ML) may harbor the potential to capture the metabolic complexity in Alzheimer’s Disease (AD). Here we set out to test the performance of metabolites in blood to categorise AD when compared to CSF biomarkers.METHODSThis study analysed samples from 242 cognitively normal (CN) people and 115 with AD-type dementia utilizing plasma metabolites (n=883). Deep Learning (DL), Extreme Gradient Boosting (XGBoost) and Random Forest (RF) were used to differentiate AD from CN. These models were internally validated using Nested Cross Validation (NCV).RESULTSOn the test data, DL produced the AUC of 0.85 (0.80-0.89), XGBoost produced 0.88 (0.86-0.89) and RF produced 0.85 (0.83-0.87). By comparison, CSF measures of amyloid, p-tau and t-tau (together with age and gender) produced with XGBoost the AUC values of 0.78, 0.83 and 0.87, respectively.DISCUSSIONThis study showed that plasma metabolites have the potential to match the AUC of well-established AD CSF biomarkers in a relatively small cohort. Further studies in independent cohorts are needed to validate whether this specific panel of blood metabolites can separate AD from controls, and how specific it is for AD as compared with other neurodegenerative disorders

Download Full-text

Independent morphological variables correlate with Aging, Mild Cognitive Impairment, and Alzheimer's Disease

10.1101/2022.01.10.22268812 ◽

2022 ◽

Author(s):

Fernanda Hansen Pacheco de Moraes ◽

Felipe Sudo ◽

Marina Monteiro Carneiro ◽

Bruno R. P. de Melo ◽

Paulo Mattos ◽

...

Keyword(s):

Alzheimer’S Disease ◽

Alzheimer's Disease ◽

Cognitive Impairment ◽

Mild Cognitive Impairment ◽

Cortical Thickness ◽

Healthy Aging ◽

Brain Morphology ◽

Elderly Subjects ◽

Csf Biomarkers ◽

Morphological Criteria

This manuscript presents a study with recruited volunteers that comprehends three sorts of events present in Alzheimer's Disease (AD) evolution (structural, biochemical, and cognitive) to propose an update in neurodegeneration biomarkers for AD. The novel variables, K, I, and S, suggested based on physics properties and empirical evidence, are defined by power-law relations between cortical thickness, exposed and total area, and natural descriptors of brain morphology. Our central hypothesis is that variable K, almost constant in healthy human subjects, is a better discriminator of a diseased brain than the current morphological biomarker, Cortical Thickness, due to its aggregated information. We extracted morphological features from 3T MRI T1w images of 123 elderly subjects: 77 Healthy Cognitive Unimpaired Controls (CTL), 33 Mild Cognitive Impairment (MCI) patients, and 13 Alzheimer's Disease (AD) patients. Moreover, Cerebrospinal Fluid (CSF) biomarkers and clinical data scores were correlated with K, intending to characterize health and disease in the cortex with morphological criteria and cognitive-behavioral profiles. K distinguishes Alzheimer's Disease, Mild Cognitive Impairment, and Healthy Cognitive Unimpaired Controls globally and locally with reasonable accuracy (CTL-AD, 0.82; CTL-MCI, 0.58). Correlations were found between global and local K associated with clinical behavioral data (executive function and memory assessments) and CSF biomarkers (t-Tau, Aβ-40, and Aβ-42). The results suggest that the cortical folding component, K, is a premature discriminator of healthy aging, Mild Cognitive Impairment, and Alzheimer's Disease, with significant differences within diagnostics. Despite the non-concomitant events, we found correlations between brain structural degeneration (K), cognitive tasks, and biochemical markers.

Download Full-text

Defining Multivariate Normative Rules for Healthy Aging using Neuroimaging and Machine Learning: An Application to Alzheimer's Disease

Journal of Alzheimer s Disease ◽

10.3233/jad-140189 ◽

2014 ◽

Vol 43 (1) ◽

pp. 201-212 ◽

Cited By ~ 7

Author(s):

Ailton Andrade de Oliveira ◽

Maria Teresa Carthery-Goulart ◽

Pedro Paulo de Magalhães Oliveira ◽

Daniel Carneiro Carrettiero ◽

João Ricardo Sato ◽

...

Keyword(s):

Machine Learning ◽

Alzheimer’S Disease ◽

Alzheimer's Disease ◽

Healthy Aging

Download Full-text

Unravelling the Association Between Amyloid-PET and Cerebrospinal Fluid Biomarkers in the Alzheimer’s Disease Spectrum: Who Really Deserves an A+?

Journal of Alzheimer s Disease ◽

10.3233/jad-210593 ◽

2021 ◽

pp. 1-12

Author(s):

Luca Sacchi ◽

Tiziana Carandini ◽

Giorgio Giulio Fumagalli ◽

Anna Margherita Pietroboni ◽

Valeria Elisa Contarino ◽

...

Keyword(s):

Alzheimer’S Disease ◽

Alzheimer's Disease ◽

Cerebrospinal Fluid ◽

Amyloid Β ◽

Standardized Uptake Value ◽

Csf Biomarkers ◽

Amyloid Pet ◽

Linear Regression Models ◽

Csf Analysis ◽

Disease Spectrum

Background: Association between cerebrospinal fluid (CSF)-amyloid-β (Aβ)42 and amyloid-PET measures is inconstant across the Alzheimer’s disease (AD) spectrum. However, they are considered interchangeable, along with Aβ 42/40 ratio, for defining ‘Alzheimer’s Disease pathologic change’ (A+). Objective: Herein, we further characterized the association between amyloid-PET and CSF biomarkers and tested their agreement in a cohort of AD spectrum patients. Methods: We include ed 23 patients who underwent amyloid-PET, MRI, and CSF analysis showing reduced levels of Aβ 42 within a 365-days interval. Thresholds used for dichotomization were: Aβ 42 < 640 pg/mL (Aβ 42+); pTau > 61 pg/mL (pTau+); and Aβ 42/40 < 0.069 (ADratio+). Amyloid-PET scans were visually assessed and processed by four pipelines (SPMCL, SPMAAL, FSGM, FSWC). Results: Different pipelines gave highly inter-correlated standardized uptake value ratios (SUVRs) (rho = 0.93–0.99). The most significant findings were: pTau positive correlation with SPMCL SUVR (rho = 0.56, p = 0.0063) and Aβ 42/40 negative correlation with SPMCL and SPMAAL SUVRs (rho = –0.56, p = 0.0058; rho = –0.52, p = 0.0117 respectively). No correlations between CSF-Aβ 42 and global SUVRs were observed. In subregion analysis, both pTau and Aβ 42/40 values significantly correlated with cingulate SUVRs from any pipeline (R2 = 0.55–0.59, p < 0.0083), with the strongest associations observed for the posterior/isthmus cingulate areas. However, only associations observed for Aβ 42/40 ratio were still significant in linear regression models. Moreover, combining pTau with Aβ 42 or using Aβ 42/40, instead of Aβ 42 alone, increased concordance with amyloid-PET status from 74% to 91% based on visual reads and from 78% to 96% based on Centiloids. Conclusion: We confirmed that, in the AD spectrum, amyloid-PET measures show a stronger association and a better agreement with CSF-Aβ 42/40 and secondarily pTau rather than Aβ 42 levels.

Download Full-text