Attention deficit hyperactivity disorder (ADHD) is one of the most common mental-health disorders. As a neurodevelopment disorder, neuroimaging technologies, such as magnetic resonance imaging (MRI), coupled with machine learning algorithms, are being increasingly explored as biomarkers in ADHD. Among various machine learning methods, deep learning has demonstrated excellent performance on many imaging tasks. With the availability of publically-available, large neuroimaging data sets for training purposes, deep learning-based automatic diagnosis of psychiatric disorders can become feasible. In this paper, we develop a deep learning-based ADHD classification method via 3-D convolutional neural networks (CNNs) applied to MRI scans. Since deep neural networks may utilize millions of parameters, even the large number of MRI samples in pooled data sets is still relatively limited if one is to learn discriminative features from the raw data. Instead, here we propose to first extract meaningful 3-D low-level features from functional MRI (fMRI) and structural MRI (sMRI) data. Furthermore, inspired by radiologists’ typical approach for examining brain images, we design a 3-D CNN model to investigate the local spatial patterns of MRI features. Finally, we discover that brain functional and structural information are complementary, and design a multi-modality CNN architecture to combine fMRI and sMRI features. Evaluations on the hold-out testing data of the ADHD-200 global competition shows that the proposed multi-modality 3-D CNN approach achieves the state-of-the-art accuracy of 69.15% and outperforms reported classifiers in the literature, even with fewer training samples. We suggest that multi-modality classification will be a promising direction to find potential neuroimaging biomarkers of neurodevelopment disorders.

During the last decades, smart tactile sensing systems based on different sensing techniques have been developed due to their high potential in industry and biomedical engineering. However, smart tactile sensing technologies and systems are still in their infancy, as many technological and system issues remain unresolved and require strong interdisciplinary efforts to address them. This paper provides an overview of smart tactile sensing systems, with a focus on signal processing technologies used to interpret the measured information from tactile sensors and/or sensors for other sensory modalities. The tactile sensing transduction and principles, fabrication and structures are also discussed with their merits and demerits. Finally, the challenges that tactile sensing technology needs to overcome are highlighted.

Advances in flexible electronic materials and smart textile, along with broad availability of smart phones, cloud and wireless systems have empowered the wearable technologies for significant impact on future of digital and personalized healthcare as well as consumer electronics. However, challenges related to lack of accuracy, reliability, high power consumption, rigid or bulky form factor and difficulty in interpretation of data have limited their wide-scale application in these potential areas. As an important solution to these challenges, we present latest advances in novel flexible electronic materials and sensors that enable comfortable and conformable body interaction and potential for invisible integration within daily apparel. Advances in novel flexible materials and sensors are described for wearable monitoring of human vital signs including, body temperature, respiratory rate and heart rate, muscle movements and activity. We then present advances in signal processing focusing on motion and noise artifact removal, data mining and aspects of sensor fusion relevant to future clinical applications of wearable technology.

Due to its remarkable feature representation capability and high performance, convolutional neural networks (CNN) have emerged as a popular choice for hyperspectral image (HSI) analysis. However, the performances of traditional CNN-based patch-wise classification methods are limited by insufficient training samples, and the evaluation strategies tend to provide overoptimistic results due to training-test information leakage. To address these concerns, we propose a novel spectral-spatial 3-D fully convolutional network (SS3FCN) to jointly explore the spectral-spatial information and the semantic information. SS3FCN takes small patches of original HSI as inputs and produces the corresponding sized outputs, which enhances the utilization rate of the scarce labeled images and boosts the classification accuracy. In addition, to avoid the potential information leakage and make a fair comparison, we introduce a new principle to generate classification benchmarks. Experimental results on four popular benchmark datasets, including Salinas Valley, Pavia University, Indian Pines, and Houston University, demonstrate that the SS3FCN outperforms state-of-the-art methods and can be served as a baseline for future research on HSI classification.

Breast cancer is a common disease in elderly women. With the development of microarray technique, discovering gene signature became a powerful approach in predicting survival of breast cancer. Previously, a 70-gene signature had been discovered for breast cancer prognosis prediction and received a good performance. In this study we adopted an efficient feature selection method: the support vector machine-based recursive feature elimination (SVM-RFE) approach for gene selection and prognosis prediction. Using the leave-one-out evaluation procedure on a gene expression dataset including 295 breast cancer patients, we discovered a 50-gene signature that by combing with SVM, achieved a superior prediction performance with 34%, 48% and 3% improvement in Accuracy, Sensitivity and Specificity, compared with the widely used 70-gene signature. Further analysis shows that the 50-gene signature is effective in predicting the prognoses of metastases and distinguishing patient who should receive adjuvant therapy.

Coal-gangue separation is an essential step in the coal preparation process. However, existing manual selection and mechanical separation methods require a large amount of labor, consume too much water, and involve health hazards. In this study, we solve this problem based on near infrared spectroscopy (NIRS) technique in tandem with improved broad learning. First, to remove outliers effectively, we propose an improved Mahalanobis distance-based method against masking and swamping effects when there is more than one outlier in the dataset. Second, we employ least absolute shrinkage and selection operator (lasso) regularization to optimize the model structure and achieve state-of-the-art accuracy of 99.01%10.0113. Furthermore, the designed spectra acquisition device is able to automatically adjust the stand-off distance to the identified optimal value and a corresponding software for coal-gangue identification is released. The developed strategy can be applied to coal/gangue blocks, not limited to the powder samples, which is an attempt to progress towards a more realistic application. The experimental results demonstrate that the proposed strategy is of great potential for nondestructive identification of gangue from coal.