CLADSI: Deep Continual Learning for Alzheimer's Disease Stage Identification Using Accelerometer Data.

Alzheimer's disease (AD) is a neurodegenerative disorder that can cause a significant impairment in physical and cognitive functions. Gait disturbances are also reported as a symptom of AD. Previous works have used Convolutional Neural Networks (CNNs) to analyze data provided by motion sensors that monitor Alzheimer's patients. However, these works have not explored continual learning algorithms that allow the CNN to configure itself as it receives new data from these sensors. This work proposes a method aimed at enabling CNNs to learn from a continuous stream of data from motion sensors without having full access to previous data. The CNN identifies the stage of AD from the analysis of data provided by motion sensors. The work includes an experimentation with data captured by accelerometers that monitored the activity of 35 Alzheimer's patients for a week in a daycare center. The CNN achieves an accuracy of 86,94%, 86,48% and 84,37% for 2, 3 and 4 experiences respectively. The proposal provides advantages to working with a continuous stream of data so that the CNN are constantly self-configuring without the intervention of a human. The work can be considered as promising and helpful in finding deep learning solutions in medical cases in which patients are constantly monitored.

Alzheimer's Disease stage identification using deep learning models.

OBJECTIVE: The aim of this research is to identify the stage of Alzheimer's Disease (AD) patients through the use of mobility data and deep learning models. This process facilitates the monitoring of the disease and allows actions to be taken in order to provide the optimal treatment and the prevention of complications. MATERIALS AND METHODS: We employed data from 35 patients with AD collected by smartphones for a week in a daycare center. The data sequences of each patient recorded the accelerometer changes while daily activities were performed and they were labeled with the stage of the disease (early, middle or late). Our methodology processes these time series and uses a Convolutional Neural Network (CNN) model to recognize the patterns that identify each stage. RESULTS: The CNN-based method achieved a 90.91% accuracy and an F1-score of 0.897, greatly improving the results obtained by the traditional feature-based classifiers. DISCUSSION AND CONCLUSION: In our research, we show that mobility data can be a valuable resource for the treatment of patients with AD as well as to study the progress of the disease. The use of our CNN-based method improves the accuracy of the identification of AD stages in comparison to common supervised learning models.

Classification of Alzheimer's Patients through Ubiquitous Computing.

Functional data analysis and artificial neural networks are the building blocks of the proposed methodology that distinguishes the movement patterns among c's patients on different stages of the disease and classifies new patients to their appropriate stage of the disease. The movement patterns are obtained by the accelerometer device of android smartphones that the patients carry while moving freely. The proposed methodology is relevant in that it is flexible on the type of data to which it is applied. To exemplify that, it is analyzed a novel real three-dimensional functional dataset where each datum is observed in a different time domain. Not only is it observed on a difference frequency but also the domain of each datum has different length. The obtained classification success rate of 83 % indicates the potential of the proposed methodology.