Unified Modeling of Imputation, Forecasting, and Prediction for AD Progression

Jung, Wonsik; Mulyadi, Ahmad Wisnu; Suk, Heung-Il

doi:10.1007/978-3-030-32251-9_19

Cited by 10 publications

(15 citation statements)

References 7 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…• Compared to the competing methods considered in our experiments, our proposed method outperformed them in all metrics for imputation, regression, and classification. This work extends the preliminary version published by (Jung et al, 2019) by revising the network architecture and performing more rigorous experiments and analyses.…”

Section: Introductionmentioning

confidence: 73%

“…More recently, we have witnessed the potential of deep learning methods (Wang et al, 2018;Lee et al, 2019;Ghazi et al, 2019;Jung et al, 2019) for DPM thanks to their favorable characteristics of learning feature representations from data, rather than engineering feature manually. Especially, recurrent neural networks (RNNs) and their variants have been mostly employed because of their methodological ability to handle the time alignment issue among intra-and inter-subject trajectories.…”

Section: Introductionmentioning

confidence: 99%

“…Especially, recurrent neural networks (RNNs) and their variants have been mostly employed because of their methodological ability to handle the time alignment issue among intra-and inter-subject trajectories. In particular, long short-term memory (LSTM) (Hochreiter and Schmidhuber, 1997) is one of the successful techniques to encode temporal patterns (Wang et al, 2018;Yoon et al, 2018;Ghazi et al, 2019;Jung et al, 2019) by capturing long-term dependencies among multivariate measurements over time with efficient gating-based operations (Gers et al, 1999;Gers and Schmidhuber, 2001). However, its use in DPM remains challenging with incomplete samples, i.e., missing measurements in the data.…”

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Deep Recurrent Model for Individualized Prediction of Alzheimer's Disease Progression

Jung¹,

Jun²,

Suk³

2020

Preprint

Self Cite

View full text Add to dashboard Cite

Alzheimer's disease (AD) is known as one of the major causes of dementia and is characterized by slow progression over several years, with no treatments or available medicines. In this regard, there have been efforts to identify the risk of developing AD in its earliest time. While many of the previous works considered crosssectional analysis, more recent studies have focused on the diagnosis and prognosis of AD with longitudinal or time series data in a way of disease progression modeling (DPM). Under the same problem settings, in this work, we propose a novel computational framework that forecasts the phenotypic measurements of MRI biomarkers and predicts the clinical statuses at multiple future time points. However, in handling time series data, it generally faces with many unexpected missing observations. In regard to such an unfavorable situation, we define a secondary problem of estimating those missing values and tackle it in a systematic way by taking account of temporal and multivariate relations inherent in time series data. Concretely, we propose a deep recurrent network that jointly tackles the three problems of (i) missing value imputation, (ii) phenotypic measurements forecasting, and (iii) clinical status prediction of a subject based on his/her longitudinal imaging biomarkers. Notably, the learnable model parameters of our network are trained in an end-to-end manner with our circumspectly defined loss function. In our experiments over The Alzheimer's Disease Prediction Of Longitudinal Evolution (TADPOLE) challenge cohort, we measured performance for various metrics and compared our method to competing methods in the literature. Exhaustive analyses and ablation studies were also conducted to better confirm the effectiveness of our method.

show abstract

Section: Introductionmentioning

confidence: 73%

Section: Introductionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Deep Recurrent Model for Individualized Prediction of Alzheimer's Disease Progression

Jung¹,

Jun²,

Suk³

2020

Preprint

Self Cite

View full text Add to dashboard Cite

show abstract

“…Current deep learning techniques for AD analysis are focused mainly on the diagnosis and prediction of structural change or cognitive scores of AD (Li and Fan, 2019;Parisot et al, 2018;Spasov et al, 2019;Zhang et al, 2017). It includes classification of the future AD stages (Basu et al, 2019) or time of conversion from one state to another (Lee et al, 2019;Lorenzi et al, 2019), and regression of biomarker values, such as cognitive scores and ventricle volumes (Ghazi et al, 2019;Jung et al, 2019). Prediction of AD stage and conversion time were mainly conducted with Recurrent Neural Networks (RNN), including Long-Short Term Memory (LSTM) networks (Ghazi et al, 2019;Lee et al, 2019;Li and Fan, 2019), in which biomarkers collected at each time go through a node of the RNN, and the output of the network in each later node is the prediction score.…”

Section: Deep Learning For Ad Longitudinal Biomarkersmentioning

confidence: 99%

DeepAtrophy: Teaching a neural network to detect progressive changes in longitudinal MRI of the hippocampal region in Alzheimer's disease

et al. 2021

View full text Add to dashboard Cite

“…The FedHome framework can also adopt other deep neural networks tailored to specific task, for example, recurrent neural networks for modeling Alzheimer's Disease (AD) progression [30], attention network for dementia status prediction from brain magnetic resonance imaging [31]. This makes FedHome as a gen-eral framework flexible for supporting many privacy-preserving healthcare applications.…”

Section: The Holistic Algorithmmentioning

confidence: 99%

FedHome: Cloud-Edge based Personalized Federated Learning for In-Home Health Monitoring

Chen

Zhou

et al. 2020

Preprint

View full text Add to dashboard Cite

In-home health monitoring has attracted great attention for the ageing population worldwide. With the abundant user health data accessed by Internet of Things (IoT) devices and recent development in machine learning, smart healthcare has seen many successful stories. However, existing approaches for in-home health monitoring do not pay sufficient attention to user data privacy and thus are far from being ready for large-scale practical deployment. In this paper, we propose FedHome, a novel cloud-edge based federated learning framework for in-home health monitoring, which learns a shared global model in the cloud from multiple homes at the network edges and achieves data privacy protection by keeping user data locally. To cope with the imbalanced and non-IID distribution inherent in user's monitoring data, we design a generative convolutional autoencoder (GCAE), which aims to achieve accurate and personalized health monitoring by refining the model with a generated class-balanced dataset from user's personal data. Besides, GCAE is lightweight to transfer between the cloud and edges, which is useful to reduce the communication cost of federated learning in FedHome. Extensive experiments based on realistic human activity recognition data traces corroborate that FedHome significantly outperforms existing widely-adopted methods.

show abstract

Unified Modeling of Imputation, Forecasting, and Prediction for AD Progression

Cited by 10 publications

References 7 publications

Deep Recurrent Model for Individualized Prediction of Alzheimer's Disease Progression

Deep Recurrent Model for Individualized Prediction of Alzheimer's Disease Progression

DeepAtrophy: Teaching a neural network to detect progressive changes in longitudinal MRI of the hippocampal region in Alzheimer's disease

FedHome: Cloud-Edge based Personalized Federated Learning for In-Home Health Monitoring

Contact Info

Product

Resources

About