A Machine Learning Framework for Predicting Dementia and Mild Cognitive Impairment

Abstract

Dementia is one of the most feared illnesses that has a growing year-to-year negative global impact, having a health and social care cost higher than cancer, stroke and chronic heart disease, taken together. Without the availability of a cure, nor a standardised clinical test, the utilisation of machine learning methods to identify individuals that are at risk of developing dementia could bring a new step towards proactive intervention. This study's goal is to carry out a precursor analysis leading to building classification models with enhanced capabilities for differentiating diagnoses of CN (Cognitively Normal), MCI (Mild Cognitive Impairment) and Dementia. The predictive modelling approach we propose is based on the ReliefF method combined with statistical permutation tests for feature selection, and on model training, tuning, and testing based on algorithms such as Random Forests, Support Vector Machines, Gaussian Processes, Stochastic Gradient Boosting, and eXtreme Gradient Boosting. Stability of model performances were studied in computationally intensive Monte Carlo simulations. The results consistently show that our models accurately detect dementia, and also mild cognitive impairment patients by only using the inclusion of baseline measurements as predictors, thus illustrating the importance of baseline measurements. The best results issued from Monte Carlo were achieved by eXtreme Gradient Boosting optimised models, with an accuracy of 0.88 (SD 0.02), a sensitivity of 0.93 (SD 0.02) and a specificity of 0.94 (SD 0.01) for dementia, and a sensitivity of 0.86 (SD 0.02) and a specificity of 0.9 (SD 0.02) for mild cognitive impairment. These results support in particular future developments for a riskbased method that can identify an individual's risk of developing dementia.

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

“…DL algorithms are known to often take advantage of large and/or unstructured data (such as images) to produce more accurate category discrimination/prediction. In a study using the Alzheimer's Disease Neuroimaging Initiative (ADNI) data for AD prediction, XGBoost demonstrated superior results (AUC = 0.97 (0.01) when including imaging parameters (MRI and PET) as predictors and when compared to RF, Support Vector Machines, Gaussian Processes and Stochastic Gradient Boosting [16]. In another study where cognition and MRI were used as predictors, Kernel Ridge Regression was performed to R 2 = 0.87 (0.025) when cognition and MRI predictors were included [17].…”

Risk Score Stratification of Alzheimer’s Disease and Mild Cognitive Impairment using Deep Learning

“…Stamate et al . identified mPACCdigit, mPACCtrailsB, LDELTOTAL as the top classifiers [10]. They combined these scores with PET and MRI data and achieved an AUC of 0.88 for the binary classification of NC versus dementia.…”

“…A large array of neurocognitive tests are currently used to detect cognitive impairment and classify amongst normal controls (CN), EMCI, LMCI and AD [7][8]. Many studies have identified a few top classifiers using logistic regression and machine learning methods [9][10][11][12][13][14][15][16][17][18]. Some studies have also used MRI and genetic data in conjunction with neurocognitive measures for classification [19][20].…”

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

“…DL algorithms are known to often take advantage of large and/or unstructured data (such as images) to produce more accurate category discrimination/ prediction. In a study using the Alzheimer’s Disease Neuroimaging Initiative (ADNI) data for AD prediction, XGBoost demonstrated superior results (AUC= 0.97 (0.01)) when including imaging parameters (MRI and PET) as predictors and when compared to RF, Support Vector Machines, Gaussian Processes and Stochastic Gradient Boosting [16]. In other study where cognition and MRI were used as predictors, Kernel Ridge Regression performed to R 2 =0.87 (0.025) when cognition and MRI predictors were included [17].…”