Breast tumor segmentation of ultrasound images provides valuable information of tumors for early detection and diagnosis. Accurate segmentation is challenging due to low image contrast between areas of interest; speckle noises, and large inter-subject variations in tumor shape and size. This paper proposes a novel Multi-scale Dynamic Fusion Network (MDF-Net) for breast ultrasound tumor segmentation. It employs a two-stage end-to-end architecture with a trunk sub-network for multiscale feature selection and a structurally optimized refinement sub-network for mitigating impairments such as noise and inter-subject variation via better feature exploration and fusion. The trunk network is extended from UNet++ with a simplified skip pathway structure to connect the features between adjacent scales. Moreover, deep supervision at all scales, instead of at the finest scale in UNet++, is proposed to extract more discriminative features and mitigate errors from speckle noise via a hybrid loss function. Unlike previous works, the first stage is linked to a loss function of the second stage so that both the preliminary segmentations and refinement subnetworks can be refined together at training. The refinement sub-network utilizes a structurally optimized MDF mechanism to integrate preliminary segmentation information (capturing general tumor shape and size) at coarse scales and explores inter-subject variation information at finer scales. Experimental results from two public datasets show that the proposed method achieves better Dice and other scores over state-of-the-art methods. Qualitative analysis also indicates that our proposed network is more robust to tumor size/shapes, speckle noise and heavy posterior shadows along tumor boundaries. An optional post-processing step is also proposed to facilitate users in mitigating segmentation artifacts. The efficiency of the proposed network is also illustrated on the "Electron Microscopy neural structures segmentation dataset". It outperforms a state-of-the-art algorithm based on UNet-2022 with simpler settings. This indicates the advantages of our MDF-Nets in other challenging image segmentation tasks with small to medium data sizes.

Ensuring the positive definiteness and avoiding ill conditioning of the Hessian update in the stochastic Broyden-Fletcher-Goldfarb-Shanno (BFGS) method are significant in solving nonconvex problems. This article proposes a novel stochastic version of a damped and regularized BFGS method for addressing the above problems. While the proposed regularized strategy helps to prevent the BFGS matrix from being close to singularity, the new damped parameter further ensures the positivity of the product of correction pairs. To alleviate the computational cost of the stochastic limited memory BFGS (LBFGS) updates and to improve its robustness, the curvature information is updated using the averaged iterate at spaced intervals. The effectiveness of the proposed method is evaluated through the logistic regression and Bayesian logistic regression problems in machine learning. Numerical experiments are conducted by using both synthetic data set and several real data sets. The results show that the proposed method generally outperforms the stochastic damped LBFGS (SdLBFGS) method. In particular, for problems with small sample sizes, our method has shown superior performance and is capable of mitigating ill-conditioned problems. Furthermore, our method is more robust to the variations of the batch size and memory size than the SdLBFGS method.

This paper proposes a novel multi-Laplacian prior (MLP) and augmented Lagrangian method (ALM) approach for gene interactions and putative transcription factors (TFs) identification from time-course gene microarray data. It employs a non-linear time-varying auto-regressive (N-TVAR) model and the Maximum-A-Posteriori-Probability method for incorporating the multi-Laplacian prior and the continuity constraint. The MLP allows connections to/from a gene to be better preserved for putative TF identification in non-stationarity gene regulatory network as compared with conventional$L_1$-based penalties. Moreover, the ALM allows the resultant non-smooth$L_1$-based penalties to be decoupled from the remaining smooth terms, so that the former and latter can be efficiently solved using a low-complexity proximity operator and smooth optimization technique, respectively. Synthetic and real time-course gene microarray datasets are tested to evaluate the performance of the proposed method. Experimental results show that the proposed method gives better accuracy and higher computational speed than our previous work using smoothed approximation. Moreover, its performance, without the use of ChIP-chip data, is found to be highly comparable with other state-of-the-art methods integrating both ChIP-chip and gene microarray data. It suggests that the proposed method may serve as a useful exploratory tool for putative TF identification with reduced experimental cost.

This paper proposes a novel consensus gene selection criteria for partial least squares-based gene microarray analysis. By quantifying the extent of consistency and distinctiveness of the differential gene expressions across different double cross validations (CV) or randomizations in terms of occurrence and randomization p-values, the proposed criteria are able to identify a more comprehensive genes associated with the underlying disease. A Distributed GPU implementation has been proposed to accelerate the gene selection problem and about 8-11 times speed up has been achieved based on the microarray datasets considered. Simulation results using various cancer gene microarray datasets show that the proposed approach is able to achieve highly comparable classification accuracy in comparing with many conventional approaches. Furthermore, enrichment analysis on the selected genes for Diffused Large B Cell Lymphoma (DLBCL) and Prostate Cancer datasets and show that only the proposed approach is able to identify gene lists enriched in different pathways with significant p-values. In contrast, sufficient statistical significance cannot be found for conventional SVM-RFE and the t-test. The reliability in identifying and establishing statistical significance of the gene findings makes the proposed approach an attractive alternative for cancer related researches based on gene expression profiling or other similar data.

This paper presents a distributed optimal power flow approach based on Augmented Lagrangian (AL) and Sequential Quadratic Programming (SQP). It is able to separate the OPF into smaller sub-problems, which could be iteratively solved individually using the SQP. This utilizes the SQP for largescale problems with non-linear objective functions and constraints. Simulation and comparison using the IEEE 30 and 118 buses examples show that the proposed distributed approach is able to achieve comparable performance with other benchmark centralized solvers provided by the FMINCON in MATPOWER. This suggests the proposed approach may serve an attractive alternative to other OPF algorithms.

Missing data imputation is often required in wireless sensor networks (WSNs) to fill up missing measurements due to transmission loss, hardware failure and other factors. In this paper, we propose new variable forgetting factor (VFF) and regularization extensions to the recursive dynamic factor analysis (RDFA) algorithm for imputation of missing data in WSN data. It takes advantage of the correlated structure of the redundancy among WSN measurements by decomposing WSN measurements into orthogonal factor loadings and de-correlated factors. A new local polynomial model (LPM) based variable forgetting factor is proposed for the RDFA algorithm and it enables us to better adapt to the time-varying environment. Finally, ℓ2 regularization is further incorporated to RDFA for improving the numerical conditioning. Experimental results using a real WSN dataset show that the proposed algorithm is able to achieve better accuracy than other conventional approaches.

This paper presents preliminary results and findings of a dynamic gene regulatory network analysis obtained from Saccharomyces cerevisiae (budding yeast) time-course DNA microarray data using a new Alternative Direction Methods of Multipliers (ADMM) based maximum a posteriori probability and time-varying autoregression model (MAP-TVAR) approach. It employs the Li-regularization based sparsity and continuity constraints, which facilitate the identification of sparse GRNs and reduce the estimation variance respectively. Simulation results using synthetic dataset show that the proposed ADMM-based extension not only performs better than our previous work in terms of identification accuracy but also is able to achieve considerable speedup. This enables us to process the whole genome of the budding yeast containing 10,715 genes and 15 timepoints more efficiently. We are able to identify gene interactions aligning well with some natural phenomena and reported in yeast cell cycle related literature. These suggest that the MAP-TVAR approach may serve as a useful tool for large-scale time-varying GRNs analysis using gene microarray data and other related datasets.

Ultrasonography is an important diagnostic imaging technique for visualization of tendons, which provides useful health diagnostic and fundamental information in neuromuscular studies of human motion systems. Conventional ultrasonic-based tendon studies, however, are highly dependent on subjective experience of operators due to various impairments of ultrasound images. Dynamic changes of muscle and tendon deformation in a sequence can hardly be manually processed. Consequently, there is an urgent need for automatic analysis of tendon behavior. This paper proposes an automatic ultrasonic tendon tracking algorithm to extract the shape deformation of central tendon of rectus femoris (CT-RF) from ultrasonic image sequences. The tracking problem is complicated by the highly deformable tendon, time-varying brightness, and the inconspicuousness of the target. To address this difficult tracking problem, we proposed a new intensity-compensated free-form deformation (IC-FFD)-based tracking algorithm with local shape refinement (LSR). Experimental results and comparison show that the proposed IC-FFD-LSR algorithm outperforms IC-FFD and conventional methods such as MI-FFD in CT-RF tracking.

Studies of time-varying or dynamic brain connectivity (BC) using functional magnetic resonance imaging (fMRI) are crucial to understand the relationship between different brain regions. This paper presents a novel method for estimating dynamic BC using a time-varying multivariate autoregressive (AR) model with spatial sparsity and temporal continuity constraints. The problem is formulated as a maximum a posterior probability (MAP) estimation problem and solved as a least square problem with Li-regularization for imposing the constraints. The Limited-memory Broyden-Fletcher-Goldfarb-Shanno (L-BFGS) method is employed to estimate the model parameters for making inference of dynamic BC. The proposed method was evaluated using synthetic data and visual checkerboard task experiment fMRI data. The results show that the method can effectively capture transient information transfer among visual-related brain regions whereas controlled areas not related to the process remain inactive. These verify the effectiveness and reduced variance of the proposed method for investigating dynamic task-related BC from fMRI data.

This paper proposes a dynamic nonlinear autoregressive model based algorithm for gene regulatory networks (GRNs) identification with biological stage change detection using the L1-regularization. This allows subtle variations in the same state to be penalized and prominent changes across adjacent states to be captured. Furthermore, by assuming local-stationarity within each detected biological state, the number of network parameters can be significantly reduced. Simulation results using a dynamic synthetic dataset and a real time course Drosophila Melanogaster DNA microarray dataset shows that the proposed method is able to achieve better identification accuracy in comparing with other conventional approaches. Moreover, it is able to identify the biological state change point precisely and identify the GRNs with effectiveness. These suggest that the proposed approach may provide an attractive alternative in GRNs identification problem.

Unlike most conventional techniques with static model assumption, this paper aims to estimate the time-varying model parameters and identify significant genes involved at different timepoints from time course gene microarray data. We first formulate the parameter identification problem as a new maximum a posteriori probability estimation problem so that prior information can be incorporated as regularization terms to reduce the large estimation variance of the high dimensional estimation problem. Under this framework, sparsity and temporal consistency of the model parameters are imposed using L1-regularization and novel continuity constraints, respectively. The resulting problem is solved using the L-BFGS method with the initial guess obtained from the partial least squares method. A novel forward validation measure is also proposed for the selection of regularization parameters, based on both forward and current prediction errors. The proposed method is evaluated using a synthetic benchmark testing data and a publicly available yeast Saccharomyces cerevisiae cell cycle microarray data. For the latter particularly, a number of significant genes identified at different timepoints are found to be biological significant according to previous findings in biological experiments. These suggest that the proposed approach may serve as a valuable tool for inferring time-varying gene regulatory networks in biological studies.

The Logistic Principal Component Regression (LPCR) has found many applications in classification of high-dimensional data, such as tumor classification using microarray data. However, when the measurements are contaminated and/or the observations are mislabeled, the performance of the LPCR will be significantly degraded. In this paper, we propose a new robust LPCR based on M-estimation, which constitutes a versatile framework to reduce the sensitivity of the estimators to outliers. In particular, robust detection rules are used to first remove the contaminated measurements and then a modified Huber function is used to further remove the contributions of the mislabeled observations. Experimental results show that the proposed method generally outperforms the conventional LPCR under the presence of outliers, while maintaining a performance comparable to that obtained under normal condition.

This paper proposes a new method for preliminary identification of gene regulatory networks (GRNs) from gene microarray cancer databased on ridge partial least squares (RPLS) with recursive feature elimination (RFE) and novel Brier and occurrence probability measures. It facilitates the preliminary identification of meaningful pathways and genes for a specific disease, rather than focusing on selecting a small set of genes for classification purposes as in conventional studies. First, RFE and a novel Brier error measure are incorporated in RPLS to reduce the estimation variance using a two-nested cross validation (CV) approach. Second, novel Brier and occurrence probability-based measures are employed in ranking genes across different CV subsamples. It helps to detect different GRNs from correlated genes which consistently appear in the ranking lists. Therefore, unlike most conventional approaches that emphasize the best classification using a small gene set, the proposed approach is able to simultaneously offer good classification accuracy and identify a more comprehensive set of genes and their associated GRNs. Experimental results on the analysis of three publicly available cancer data sets, namely leukemia, colon, and prostate, show that very stable gene sets from different but relevant GRNs can be identified, and most of them are found to be of biological significance according to previous findings in biological experiments. These suggest that the proposed approach may serve as a useful tool for preliminary identification of genes and their associated GRNs of a particular disease for further biological studies using microarray or similar data.