Alzheimer’s disease (AD) is an irreversible brain degenerative disorder affecting people aged older than 65 years. Currently, there is no effective cure for AD, but its progression can be delayed with some treatments. Accurate and early diagnosis of AD is vital for the patient care and development of future treatment. Fluorodeoxyglucose positrons emission tomography (FDG-PET) is a functional molecular imaging modality, which proves to be powerful to help understand the anatomical and neural changes of brain related to AD. Most existing methods extract the handcrafted features from images, and then design a classifier to distinguish AD from other groups. These methods highly depends on the preprocessing of brain images, including image rigid registration and segmentation. Motivated by the success of deep learning in image classification, this paper proposes a new classification framework based on combination of 2D convolutional neural networks (CNN) and recurrent neural networks (RNNs), which learns the intra-slice and inter-slice features for classification after decomposition of the 3D PET image into a sequence of 2D slices. The 2D CNNs are built to capture the features of image slices while the gated recurrent unit (GRU) of RNN is cascaded to learn and integrate the inter-slice features for image classification. No rigid registration and segmentation are required for PET images. Our method is evaluated on the baseline FDG-PET images acquired from 339 subjects including 93 AD patients, 146 mild cognitive impairments (MCI) and 100 normal controls (NC) from Alzheimer’s Disease Neuroimaging Initiative (ADNI) database. Experimental results show that the proposed method achieves an area under receiver operating characteristic curve (AUC) of 95.3% for AD vs. NC classification and 83.9% for MCI vs. NC classification, demonstrating the promising classification performance.

A novel adaptive temporal-frequency network (ATFN), which is an end-to-end hybrid model incorporating deep learning networks and frequency patterns, is proposed for mid- and long-term time series forecasting. Within the framework of the ATFN, an augmented sequence to sequence model is used to learn the trend feature of complicated nonstationary time series, a frequency-domain block is used to capture dynamic and complicated periodic patterns of time series data, and a fully connected neural network is used to combine the trend and periodic features for producing a final forecast. An adaptive frequency mechanism consisting of phase adaption, frequency adaption, and amplitude adaption is designed for mapping the frequency spectrum of the current sliding window to that of the forecasting interval. The multilayer neural networks conduct a transformation similar to the inverse discrete Fourier transform for generating a periodic feature forecast. Synthetic data and real-world data with different periodic characteristics are used to evaluate the effectiveness of the proposed model. The experimental results indicate that the ATFN has promising performance and strong adaptability for long-term time-series forecasting.

Freezing of gait (FoG) prediction, combined with rhythmic laser cues, may help Parkinson’s disease (PD) patients overcome FoG episodes. This study aimed to utilize the impaired gait patterns preceding FoG to build a machine-learning-based model for FoG prediction. Acceleration signals were collected using an accelerometer attached to the lower back of 12 PD patients with FoG while they were performing designed FoG-provoking walking tasks. Step-based impaired gait features and conventional FoG detection features were extracted from the signals, based on which two FoG prediction models were built using AdaBoost to validate if the use of the impaired gait features can better predict FoG. For the correct labeling of the gait prior to FoG (pre-FoG), the personalized pre-FoG phase was defined according to the slope of the impaired gait pattern. The impaired gait features were relabeled based on the pre-FoG phase upon which the personalized labeled FoG prediction model was built. This was compared with the model built using unified labeling. Results showed that impaired gait features could better predict FoG than conventional FoG detection features with low time latency, and personalized labeling could further improve the FoG prediction accuracy. Using impaired gait features and personalized labeling, we built a FoG prediction model with 0.93 sec of latency. Compared to using conventional features and unified labeling, our model achieved 5.7% higher accuracy (82.7%) in patient-dependent test and 9.8% higher accuracy (77.9%) in patient-independent test. Therefore, using the impaired gait patterns is a promising approach to accurately predict FoG with low latency.