Compared with the rapidly growing artificial intelligence (AI) research in other branches of healthcare, the pace of developing AI capacities in dental care is relatively slow. Dental care automation, especially the automated capability for dental cone beam computed tomography (CBCT) segmentation and lesion detection, is highly needed. CBCT is an important imaging modality that is experiencing ever-growing utilization in various dental specialties. However, little research has been done for segmenting different structures, restorative materials, and lesions using deep learning. This is due to multifold challenges such as content-rich oral cavity and significant within-label variation on each CBCT image as well as the inherent difficulty of obtaining many high-quality labeled images for training. On the other hand, oral-anatomical knowledge exists in dentistry, which shall be leveraged and integrated into the deep learning design. In this article, we propose a novel anatomically constrained Dense U-Net for integrating oral-anatomical knowledge with data-driven Dense U-Net. The proposed algorithm is formulated as a regularized or constrained optimization and solved using mean-field variational approximation to achieve computational efficiency. Mathematical encoding for transforming descriptive knowledge into a quantitative form is also proposed. Our experiment demonstrates that the proposed algorithm outperforms the standard Dense U-Net in both lesion detection accuracy and dice coefficient (DICE) indices in multilabel segmentation. Benefited from the integration with anatomical domain knowledge, our algorithm performs well with data from a small number of patients included in the training.

Generally, in an advanced manufacturing system hundreds of sensors are deployed to measure key process variables in real time. Thus it is desirable to develop methodologies to use real-time sensor data for on-line system condition monitoring and anomaly detection. However, there are several challenges in developing an effective process monitoring system: (i) data streams generated by multiple sensors are high-dimensional profiles; (ii) sensor signals are affected by noise due to system-inherent variations; (iii) signals of different sensors have cluster-wise features; and (iv) an anomaly may cause only sparse changes of sensor signals. To address these challenges, this article presents a real-time multiple profiles sensor-based process monitoring system, which includes the following modules: (i) preprocessing sensor signals to remove inherent variations and conduct profile alignments, (ii) using multichannel functional principal component analysis (MFPCA)–based methods to extract sensor features by considering cluster-wise between-sensor correlations, and (iii) constructing a monitoring scheme with the top-R strategy based on the extracted features, which has scalable detection power for different fault patterns. Finally, we implement and demonstrate the proposed framework using data from a real manufacturing system.

Although several works have been proposed for multi-channel profile monitoring, two additional challenges are yet to be addressed: (i) how to model complex correlations of multi-channel profiles when different profiles have different features (i.e., weakly or sparsely correlated); (ii) how to efficiently detect sparse changes occurring in only a small segment of a few profiles. To fill this research gap, our contributions are twofold. First, we propose a novel Sparse Multi-channel Functional Principal Component Analysis (SMFPCA) to model multi-channel profiles. SMFPCA can not only flexibly describe the correlation structure of multiple, or even high-dimensional, profiles with distinct features, but also achieve sparse PCA scores which are easily interpretable. Second, we propose an efficient convergence-guaranteed optimization algorithm to solve SMFPCA in real time based on the block coordinate descent algorithm. Third, as the SMFPCA scores can naturally identify sparse out-of-control (OC) patterns, we use the scores to construct a monitoring scheme which provides increased sensitivity to sparse OC changes. Numerical studies together with a real case study in a manufacturing system demonstrate the effectiveness of the developed methodology.

Spatiotemporal data are very common in many applications, such as manufacturing systems and transportation systems. Given the intrinsic complex spatial and temporal correlations of such data, short-term and long-term prediction for spatiotemporal data is often very challenging. Most of the traditional statistical models fail to preserve innate features in data alongside their complex correlations. In this paper, we focus on a tensor-based prediction method and propose several practical techniques to improve both long-term and short-term prediction accuracy. For long-term prediction, we propose the “tensor decomposition + 2-Dimensional Auto-Regressive Moving Average (2D-ARMA)” model, and an effective way to update prediction in real-time; For short-term prediction, we propose to conduct tensor completion based on tensor clustering to avoid oversimplification and ensure accuracy. A case study based on the metro passenger flow data is conducted to demonstrate the improved performance.

The use of video-imaging data for in-line process monitoring applications has become popular in industry. In this framework, spatio-temporal statistical process monitoring methods are needed to capture the relevant information content and signal possible out-of-control states. Video-imaging data are characterized by a spatio-temporal variability structure that depends on the underlying phenomenon, and typical out-of-control patterns are related to events that are localized both in time and space. In this article, we propose an integrated spatio-temporal decomposition and regression approach for anomaly detection in video-imaging data. Out-of-control events are typically sparse, spatially clustered and temporally consistent. The goal is not only to detect the anomaly as quickly as possible (“when”) but also to locate it in space (“where”). The proposed approach works by decomposing the original spatio-temporal data into random natural events, sparse spatially clustered and temporally consistent anomalous events, and random noise. Recursive estimation procedures for spatio-temporal regression are presented to enable the real-time implementation of the proposed methodology. Finally, a likelihood ratio test procedure is proposed to detect when and where the anomaly happens. The proposed approach was applied to the analysis of high-sped video-imaging data to detect and locate local hot-spots during a metal additive manufacturing process.

Multivariate functional data from a complex system are naturally high-dimensional and have a complex cross-correlation structure. The complexity of data structure can be observed as that (1) some functions are strongly correlated with similar features, while some others may have almost no cross-correlations with quite diverse features; and (2) the cross-correlation structure may also change over time due to the system evolution. With this regard, this paper presents a dynamic subspace learning method for multivariate functional data modeling. In particular, we consider that different functions come from different subspaces, and only functions of the same subspace have cross-correlations with each other. The subspaces can be automatically formulated and learned by reformatting the problem as a sparse regression. By allowing but regularizing the regression change over time, we can describe the cross-correlation dynamics. The model can be efficiently estimated by the fast iterative shrinkage-thresholding algorithm, and the features of each subspace can be extracted using the smooth multi-channel functional principal component analysis. Some theoretical properties of the model are presented. Numerical studies, together with case studies, demonstrate the efficiency and applicability of the proposed methodology.