Cargando…

Predicting Alzheimer’s disease progression using deep recurrent neural networks()

Early identification of individuals at risk of developing Alzheimer’s disease (AD) dementia is important for developing disease-modifying therapies. In this study, given multimodal AD markers and clinical diagnosis of an individual from one or more timepoints, we seek to predict the clinical diagnos...

Descripción completa

Detalles Bibliográficos
Autores principales: Nguyen, Minh, He, Tong, An, Lijun, Alexander, Daniel C., Feng, Jiashi, Yeo, B.T. Thomas
Formato: Online Artículo Texto
Lenguaje:English
Publicado: 2020
Materias:
Acceso en línea:https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7797176/
https://www.ncbi.nlm.nih.gov/pubmed/32763427
http://dx.doi.org/10.1016/j.neuroimage.2020.117203
_version_ 1783634818136801280
author Nguyen, Minh
He, Tong
An, Lijun
Alexander, Daniel C.
Feng, Jiashi
Yeo, B.T. Thomas
author_facet Nguyen, Minh
He, Tong
An, Lijun
Alexander, Daniel C.
Feng, Jiashi
Yeo, B.T. Thomas
author_sort Nguyen, Minh
collection PubMed
description Early identification of individuals at risk of developing Alzheimer’s disease (AD) dementia is important for developing disease-modifying therapies. In this study, given multimodal AD markers and clinical diagnosis of an individual from one or more timepoints, we seek to predict the clinical diagnosis, cognition and ventricular volume of the individual for every month (indefinitely) into the future. We proposed and applied a minimal recurrent neural network (minimalRNN) model to data from The Alzheimer’s Disease Prediction Of Longitudinal Evolution (TADPOLE) challenge, comprising longitudinal data of 1677 participants (Marinescu et al., 2018) from the Alzheimer’s Disease Neuroimaging Initiative (ADNI). We compared the performance of the minimalRNN model and four baseline algorithms up to 6 years into the future. Most previous work on predicting AD progression ignore the issue of missing data, which is a prevalent issue in longitudinal data. Here, we explored three different strategies to handle missing data. Two of the strategies treated the missing data as a “preprocessing” issue, by imputing the missing data using the previous timepoint (“forward filling”) or linear interpolation (“linear filling). The third strategy utilized the minimalRNN model itself to fill in the missing data both during training and testing (“model filling”). Our analyses suggest that the minimalRNN with “model filling” compared favorably with baseline algorithms, including support vector machine/regression, linear state space (LSS) model, and long short-term memory (LSTM) model. Importantly, although the training procedure utilized longitudinal data, we found that the trained minimalRNN model exhibited similar performance, when using only 1 input timepoint or 4 input timepoints, suggesting that our approach might work well with just cross-sectional data. An earlier version of our approach was ranked 5th (out of 53 entries) in the TADPOLE challenge in 2019. The current approach is ranked 2nd out of 63 entries as of June 3rd, 2020.
format Online
Article
Text
id pubmed-7797176
institution National Center for Biotechnology Information
language English
publishDate 2020
record_format MEDLINE/PubMed
spelling pubmed-77971762021-01-10 Predicting Alzheimer’s disease progression using deep recurrent neural networks() Nguyen, Minh He, Tong An, Lijun Alexander, Daniel C. Feng, Jiashi Yeo, B.T. Thomas Neuroimage Article Early identification of individuals at risk of developing Alzheimer’s disease (AD) dementia is important for developing disease-modifying therapies. In this study, given multimodal AD markers and clinical diagnosis of an individual from one or more timepoints, we seek to predict the clinical diagnosis, cognition and ventricular volume of the individual for every month (indefinitely) into the future. We proposed and applied a minimal recurrent neural network (minimalRNN) model to data from The Alzheimer’s Disease Prediction Of Longitudinal Evolution (TADPOLE) challenge, comprising longitudinal data of 1677 participants (Marinescu et al., 2018) from the Alzheimer’s Disease Neuroimaging Initiative (ADNI). We compared the performance of the minimalRNN model and four baseline algorithms up to 6 years into the future. Most previous work on predicting AD progression ignore the issue of missing data, which is a prevalent issue in longitudinal data. Here, we explored three different strategies to handle missing data. Two of the strategies treated the missing data as a “preprocessing” issue, by imputing the missing data using the previous timepoint (“forward filling”) or linear interpolation (“linear filling). The third strategy utilized the minimalRNN model itself to fill in the missing data both during training and testing (“model filling”). Our analyses suggest that the minimalRNN with “model filling” compared favorably with baseline algorithms, including support vector machine/regression, linear state space (LSS) model, and long short-term memory (LSTM) model. Importantly, although the training procedure utilized longitudinal data, we found that the trained minimalRNN model exhibited similar performance, when using only 1 input timepoint or 4 input timepoints, suggesting that our approach might work well with just cross-sectional data. An earlier version of our approach was ranked 5th (out of 53 entries) in the TADPOLE challenge in 2019. The current approach is ranked 2nd out of 63 entries as of June 3rd, 2020. 2020-08-04 2020-11-15 /pmc/articles/PMC7797176/ /pubmed/32763427 http://dx.doi.org/10.1016/j.neuroimage.2020.117203 Text en This is an open access article under the CC BY license. (http://creativecommons.org/licenses/by/4.0/)
spellingShingle Article
Nguyen, Minh
He, Tong
An, Lijun
Alexander, Daniel C.
Feng, Jiashi
Yeo, B.T. Thomas
Predicting Alzheimer’s disease progression using deep recurrent neural networks()
title Predicting Alzheimer’s disease progression using deep recurrent neural networks()
title_full Predicting Alzheimer’s disease progression using deep recurrent neural networks()
title_fullStr Predicting Alzheimer’s disease progression using deep recurrent neural networks()
title_full_unstemmed Predicting Alzheimer’s disease progression using deep recurrent neural networks()
title_short Predicting Alzheimer’s disease progression using deep recurrent neural networks()
title_sort predicting alzheimer’s disease progression using deep recurrent neural networks()
topic Article
url https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7797176/
https://www.ncbi.nlm.nih.gov/pubmed/32763427
http://dx.doi.org/10.1016/j.neuroimage.2020.117203
work_keys_str_mv AT nguyenminh predictingalzheimersdiseaseprogressionusingdeeprecurrentneuralnetworks
AT hetong predictingalzheimersdiseaseprogressionusingdeeprecurrentneuralnetworks
AT anlijun predictingalzheimersdiseaseprogressionusingdeeprecurrentneuralnetworks
AT alexanderdanielc predictingalzheimersdiseaseprogressionusingdeeprecurrentneuralnetworks
AT fengjiashi predictingalzheimersdiseaseprogressionusingdeeprecurrentneuralnetworks
AT yeobtthomas predictingalzheimersdiseaseprogressionusingdeeprecurrentneuralnetworks