Modeling and prediction of clinical symptom trajectories in Alzheimer’s disease using longitudinal data

Abstract

Computational models predicting symptomatic progression at the individual level can be highly beneficial for early intervention and treatment planning for Alzheimer’s disease (AD). Individual prognosis is complicated by many factors including the definition of the prediction objective itself. In this work, we present a computational framework comprising machine-learning techniques for 1) modeling symptom trajectories and 2) prediction of symptom trajectories using multimodal and longitudinal data. We perform primary analyses on three cohorts from Alzheimer’s Disease Neuroimaging Initiative (ADNI), and a replication analysis using subjects from Australian Imaging, Biomarker & Lifestyle Flagship Study of Ageing (AIBL). We model the prototypical symptom trajectory classes using clinical assessment scores from mini-mental state exam (MMSE) and Alzheimer’s Disease Assessment Scale (ADAS-13) at nine timepoints spanned over six years based on a hierarchical clustering approach. Subsequently we predict these trajectory classes for a given subject using magnetic resonance (MR) imaging, genetic, and clinical variables from two timepoints (baseline + follow-up). For prediction, we present a longitudinal Siamese neural-network (LSN) with novel architectural modules for combining multimodal data from two timepoints. The trajectory modeling yields two (stable and decline) and three (stable, slow-decline, fast-decline) trajectory classes for MMSE and ADAS-13 assessments, respectively. For the predictive tasks, LSN offers highly accurate performance with 0.900 accuracy and 0.968 AUC for binary MMSE task and 0.760 accuracy for 3-way ADAS-13 task on ADNI datasets, as well as, 0.724 accuracy and 0.883 AUC for binary MMSE task on replication AIBL dataset.

Predicting sporadic Alzheimer's disease progression via inherited Alzheimer's disease‐informed machine‐learning

“…Previous studies evaluating biomarker combinations 9,12 or machine learning demonstrated that biomarker combinations can enhance the accuracy of predicting cognitive decline or conversion to dementia 20,21 . Consistently, we found increasing prediction accuracy as model estimation was informed by more biomarker modalities.…”

“…Machine learning, which is instrumental for data mining, 17 is well suited to identify biomarker sets for predicting cognitive decline. Here, recent machine-learning approaches to AD biomarker data showed that sets of imaging and CSF-derived markers discriminated MCI and AD patients from healthy controls 18 and predicted the rate of future cognitive decline 19,20 and time to symptom onset. 21 A limitation of machine learning is, however, that algorithms may perform well in the sample they were trained on but rarely generalize to new data.…”

Predicting future cognitive decline from non-brain and multimodal brain imaging data in healthy and pathological aging

“…When predicting future cognitive change, however, brain imaging might aid a prognosis with greater clinical utility that cannot be easily obtained otherwise. Most previous studies that predicted future change restricted their analysis to whether patients with mild cognitive impairment (MCI) converted to Alzheimer’s disease (AD) (Davatzikos et al, 2011; e.g., Eskildsen et al, 2015; Gaser et al, 2013; Korolev et al, 2016) or predicted membership in data-driven trajectory-groups of future decline (Bhagwat et al, 2018). Predicting future cognitive decline on a continuum (instead of forming distinct diagnostic labels from cognitive data) better characterizes the underlying change in abilities on an individual level.…”

3D-SiameseNet to Analyze Brain MRI

“…al. [2], subjects from the ADNI1 cohort were used for trajectory modelling, dividing the dataset between "Stable" and "Decline" subjects, based on the Mini Mental State Evaluation clinical scores. We selected a subset of these subjects in our dataset: 134 "Stable" subjects, and 113 "Decline" subjects.…”

“…al. [2] who have proposed a siamese network to predict the evolution of patients by using measures computed from MRI images and a cognitive score. On one hand, as we would like to take into account 3D MRI, our network is made of several layers of convolutional layers for each of its branches.…”