Dataset Information

Predicting progression from mild cognitive impairment to Alzheimer's disease on an individual subject basis by applying the CARE index across different independent cohorts.

ABSTRACT: The purposes of this study are to investigate whether the Characterizing Alzheimer's disease Risk Events (CARE) index can accurately predict progression from mild cognitive impairment (MCI) to Alzheimer's disease (AD) on an individual subject basis, and to investigate whether this model can be generalized to an independent cohort. Using an event-based probabilistic model approach to integrate widely available biomarkers from behavioral data and brain structural and functional imaging, we calculated the CARE index. We then applied the CARE index to identify which MCI individuals from the ADNI dataset progressed to AD during a three-year follow-up period. Subsequently, the CARE index was generalized to the prediction of MCI individuals from an independent Nanjing Aging and Dementia Study (NADS) dataset during the same time period. The CARE index achieved high prediction performance with 80.4% accuracy, 75% sensitivity, 82% specificity, and 0.809 area under the receiver operating characteristic (ROC) curve (AUC) on MCI subjects from the ADNI dataset over three years, and a highly validated prediction performance with 87.5% accuracy, 81% sensitivity, 90% specificity, and 0.861 AUC on MCI subjects from the NADS dataset. In conclusion, the CARE index is highly accurate, sufficiently robust, and generalized for predicting which MCI individuals will develop AD over a three-year period. This suggests that the CARE index can be usefully applied to select individuals with MCI for clinical trials and to identify which individuals will convert from MCI to AD for administration of early disease-modifying treatment.

SUBMITTER: Chen J

PROVIDER: S-EPMC6520016 | biostudies-literature | 2019 Apr

REPOSITORIES: biostudies-literature

ACCESS DATA

Publications

Predicting progression from mild cognitive impairment to Alzheimer's disease on an individual subject basis by applying the CARE index across different independent cohorts.

Chen Jiu J Chen Gang G Shu Hao H Chen Guangyu G Ward B Douglas BD Wang Zan Z Liu Duan D Antuono Piero G PG Li Shi-Jiang SJ Zhang Zhijun Z

Aging 20190401 8

The purposes of this study are to investigate whether the Characterizing Alzheimer's disease Risk Events (CARE) index can accurately predict progression from mild cognitive impairment (MCI) to Alzheimer's disease (AD) on an individual subject basis, and to investigate whether this model can be generalized to an independent cohort. Using an event-based probabilistic model approach to integrate widely available biomarkers from behavioral data and brain structural and functional imaging, we calcula ...[more]

PMID: 31078129

Similar Datasets

Project description:BackgroundMild cognitive impairment is often a precursor to dementia due to Alzheimer's disease, but many patients with mild cognitive impairment never develop dementia. New diagnostic criteria may lead to more patients receiving a diagnosis of mild cognitive impairment.ObjectiveTo develop a prediction index for the 3-year risk of progression from mild cognitive impairment to dementia relying only on information that can be readily obtained in most clinical settings.Design and participants382 participants diagnosed with amnestic mild cognitive impairment enrolled in the Alzheimer's Disease Neuroimaging Initiative (ADNI), a multi-site, longitudinal, observational study.Main predictors measuresDemographics, comorbid conditions, caregiver report of participant symptoms and function, and participant performance on individual items from basic neuropsychological scales.Main outcome measureProgression to probable Alzheimer's disease.Key resultsSubjects had a mean (SD) age of 75 (7) years and 43% progressed to probable Alzheimer's disease within 3 years. Important independent predictors of progression included being female, resisting help, becoming upset when separated from caregiver, difficulty shopping alone, forgetting appointments, number of words recalled from a 10-word list, orientation and difficulty drawing a clock. The final point score could range from 0 to 16 (mean [SD]: 4.2 [2.9]). The optimism-corrected Harrell's c-statistic was 0.71(95% CI: 0.68-0.75). Fourteen percent of subjects with low risk scores (0-2 points, n?=?124) converted to probable Alzheimer's disease over 3 years, compared to 51% of those with moderate risk scores (3-8 points, n?=?223) and 91% of those with high risk scores (9-16 points, n?=?35).ConclusionsAn index using factors that can be obtained in most clinical settings can predict progression from amnestic mild cognitive impairment to probable Alzheimer's disease and may help clinicians differentiate between mild cognitive impairment patients at low vs. high risk of progression.

Project description:Random Survival Forests (RSF) has recently showed better performance than statistical survival methods as Cox proportional hazard (CPH) in predicting conversion risk from mild cognitive impairment (MCI) to Alzheimer's disease (AD). However, RSF application in real-world clinical setting is still limited due to its black-box nature.For this reason, we aimed at providing a comprehensive study of RSF explainability with SHapley Additive exPlanations (SHAP) on biomarkers of stable and progressive patients (sMCI and pMCI) from Alzheimer's Disease Neuroimaging Initiative. We evaluated three global explanations-RSF feature importance, permutation importance and SHAP importance-and we quantitatively compared them with Rank-Biased Overlap (RBO). Moreover, we assessed whether multicollinearity among variables may perturb SHAP outcome. Lastly, we stratified pMCI test patients in high, medium and low risk grade, to investigate individual SHAP explanation of one pMCI patient per risk group.We confirmed that RSF had higher accuracy (0.890) than CPH (0.819), and its stability and robustness was demonstrated by high overlap (RBO > 90%) between feature rankings within first eight features. SHAP local explanations with and without correlated variables had no substantial difference, showing that multicollinearity did not alter the model. FDG, ABETA42 and HCI were the first important features in global explanations, with the highest contribution also in local explanation. FAQ, mPACCdigit, mPACCtrailsB and RAVLT immediate had the highest influence among all clinical and neuropsychological assessments in increasing progression risk, as particularly evident in pMCI patients' individual explanation. In conclusion, our findings suggest that RSF represents a useful tool to support clinicians in estimating conversion-to-AD risk and that SHAP explainer boosts its clinical utility with intelligible and interpretable individual outcomes that highlights key features associated with AD prognosis.

Dataset Information

Predicting progression from mild cognitive impairment to Alzheimer's disease on an individual subject basis by applying the CARE index across different independent cohorts.

Publications

Predicting progression from mild cognitive impairment to Alzheimer's disease on an individual subject basis by applying the CARE index across different independent cohorts.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets