ABSTRACT
The fusion of magnetic resonance imaging and positron emission tomography can combine biological anatomical information and physiological metabolic information, which is of great significance for the clinical diagnosis and localization of lesions. In this paper, we propose a novel adaptive linear fusion method for multi-dimensional features of brain magnetic resonance and positron emission tomography images based on a convolutional neural network, termed as MdAFuse. First, in the feature extraction stage, three-dimensional feature extraction modules are constructed to extract coarse, fine, and multi-scale information features from the source image. Second, at the fusion stage, the affine mapping function of multi-dimensional features is established to maintain a constant geometric relationship between the features, which can effectively utilize structural information from a feature map to achieve a better reconstruction effect. Furthermore, our MdAFuse comprises a key feature visualization enhancement algorithm designed to observe the dynamic growth of brain lesions, which can facilitate the early diagnosis and treatment of brain tumors. Extensive experimental results demonstrate that our method is superior to existing fusion methods in terms of visual perception and nine kinds of objective image fusion metrics. Specifically, in the results of MR-PET fusion, the SSIM (Structural Similarity) and VIF (Visual Information Fidelity) metrics show improvements of 5.61% and 13.76%, respectively, compared to the current state-of-the-art algorithm. Our project is publicly available at: https://github.com/22385wjy/MdAFuse.
Subject(s)
Algorithms , Brain Neoplasms , Brain , Magnetic Resonance Imaging , Positron-Emission Tomography , Brain Neoplasms/diagnostic imaging , Humans , Magnetic Resonance Imaging/methods , Positron-Emission Tomography/methods , Brain/diagnostic imaging , Multimodal Imaging/methods , Neural Networks, ComputerABSTRACT
In modern precision medicine, it is an important research topic to predict cancer drug response. Due to incomplete chemical structures and complex gene features, however, it is an ongoing work to design efficient data-driven methods for predicting drug response. Moreover, since the clinical data cannot be easily obtained all at once, the data-driven methods may require relearning when new data are available, resulting in increased time consumption and cost. To address these issues, an incremental broad Transformer network (iBT-Net) is proposed for cancer drug response prediction. Different from the gene expression features learning from cancer cell lines, structural features are further extracted from drugs by Transformer. Broad learning system is then designed to integrate the learned gene features and structural features of drugs to predict the response. With the capability of incremental learning, the proposed method can further use new data to improve its prediction performance without retraining totally. Experiments and comparison studies demonstrate the effectiveness and superiority of iBT-Net under different experimental configurations and continuous data learning.
Subject(s)
Antineoplastic Agents , Neoplasms , Humans , Neoplasms/drug therapy , Neoplasms/genetics , Antineoplastic Agents/pharmacology , Antineoplastic Agents/therapeutic use , Cell Line , Education, Continuing , Precision MedicineABSTRACT
This paper develops a Neural Network (NN) event-triggered finite-time consensus control method for uncertain nonlinear Multi-Agent Systems (MASs) with dead-zone input and actuator failures. In practical applications, actuator failures would inevitably arise in MASs. And the time, pattern, and value of the failures are unknown. Besides, the actuators of MASs also suffer from dead-zone nonlinearity. No matter actuator failures or dead-zone input would dramatically affect the performance and stability of MASs. To address these issues, finite-time adaptive controllers capable of simultaneously compensating for actuator failures and dead-zone input are constructed by adopting the backstepping technology. Meanwhile, the NN control scheme is adopted to handle the unknown nonlinear dynamics of each agent. Furthermore, an event-triggered control mechanism is established that no longer requires continuous communication on the control network. Under the proposed control method, all followers achieve finite-time synchronization, irrespective of the presence of limited bandwidth, unknown failures, and dead-zone input. These results are demonstrated by simulations.
ABSTRACT
Emotion recognition based on text-audio modalities is the core technology for transforming a graphical user interface into a voice user interface, and it plays a vital role in natural human-computer interaction systems. Currently, mainstream multimodal learning research has designed various fusion strategies to learn intermodality interactions but hardly considers that not all modalities play equal roles in emotion recognition. Therefore, the main challenge in multimodal emotion recognition is how to implement effective fusion algorithms based on the auxiliary structure. To address this problem, this article proposes an adaptive interactive attention network (AIA-Net). In AIA-Net, text is treated as a primary modality, and audio is an auxiliary modality. AIA-Net adapts to textual and acoustic features with different dimensions and learns their dynamic interactive relations in a more flexible way. The interactive relations are encoded as interactive attention weights to focus on the acoustic features that are effective for textual emotional representations. AIA-Net performs well in adaptively assisting the textual emotional representation with the acoustic emotional information. Moreover, multiple collaborative learning (co-learning) layers of AIA-Net achieve multiple multimodal interactions and the deep bottom-up evolution of emotional representations. Experimental results on three benchmark datasets demonstrate the great effectiveness of the proposed method over the state-of-the-art methods.
Subject(s)
Acoustics , Algorithms , Humans , Computer Systems , Emotions , LearningABSTRACT
Constrained autonomous vehicle overtaking trajectories are usually difficult to generate due to certain practical requirements and complex environmental limitations. This problem becomes more challenging when multiple contradicting objectives are required to be optimized and the on-road objects to be overtaken are irregularly placed. In this article, a novel swarm intelligence-based algorithm is proposed for producing the multiobjective optimal overtaking trajectory of autonomous ground vehicles. The proposed method solves a multiobjective optimal control model in order to optimize the maneuver time duration, the trajectory smoothness, and the vehicle visibility, while taking into account different types of mission-dependent constraints. However, one problem that could have an impact on the optimization process is the selection of algorithm control parameters. To desensitize the negative influence, a novel fuzzy adaptive strategy is proposed and embedded in the algorithm framework. This allows the optimization process to dynamically balance the local exploitation and global exploration, thereby exploring the tradeoff between objectives more effectively. The performance of using the designed fuzzy adaptive multiobjective method is analyzed and validated by executing a number of simulation studies. The results confirm the effectiveness of applying the proposed algorithm to produce multiobjective optimal overtaking trajectories for autonomous ground vehicles. Moreover, the comparison to other state-of-the-art multiobjective optimization schemes shows that the designed strategy tends to be more capable in terms of producing a set of widespread and high-quality Pareto-optimal solutions.
ABSTRACT
In this article, the problem of event-triggered tracking control for a class of uncertain nonlinear systems with unknown Prandtl-Ishlinskii (PI) hysteresis is investigated. To solve this problem, two control schemes are proposed via synthesizing the techniques of the event-triggered strategy, fuzzy-logic systems (FLSs), and adaptive backstepping control. The first basic design scheme applies an effective method to keep a balance between communication constraints and system performance under the influence of actuator PI hysteresis, while the Zeno behavior can be avoided. Furthermore, the basic design scheme not only guarantees the tracking error asymptotically converges to zero but also establishes a preserved transient performance. Nevertheless, note that the inclusive sign functions of the basic design scheme will cause possible chattering phenomenon, an alternative event-triggered adaptive control approach is then proposed. Unlike the previous control scheme, the second chattering-avoidance design approach ensures asymptotic convergence of the tracking error within a prescribed boundary δ , and finally the [Formula: see text]-norm transient performance of the tracking error is constructed. Simulations verify the established theoretical results that the proposed schemes successfully overcome the communication constraints and compensate the actuator PI hysteresis, and also present different tracking performances between two control schemes for comparison.
ABSTRACT
A prevailing problem in many machine learning tasks is that the training (i.e., source domain) and test data (i.e., target domain) have different distribution [i.e., non-independent identical distribution (i.i.d.)]. Unsupervised domain adaptation (UDA) was proposed to learn the unlabeled target data by leveraging the labeled source data. In this article, we propose a guide subspace learning (GSL) method for UDA, in which an invariant, discriminative, and domain-agnostic subspace is learned by three guidance terms through a two-stage progressive training strategy. First, the subspace-guided term reduces the discrepancy between the domains by moving the source closer to the target subspace. Second, the data-guided term uses the coupled projections to map both domains to a unified subspace, where each target sample can be represented by the source samples with a low-rank coefficient matrix that can preserve the global structure of data. In this way, the data from both domains can be well interlaced and the domain-invariant features can be obtained. Third, for improving the discrimination of the subspaces, the label-guided term is constructed for prediction based on source labels and pseudo-target labels. To further improve the model tolerance to label noise, a label relaxation matrix is introduced. For the solver, a two-stage learning strategy with teacher teaches and student feedbacks mode is proposed to obtain the discriminative domain-agnostic subspace. In addition, for handling nonlinear domain shift, a nonlinear GSL (NGSL) framework is formulated with kernel embedding, such that the unified subspace is imposed with nonlinearity. Experiments on various cross-domain visual benchmark databases show that our methods outperform many state-of-the-art UDA methods. The source code is available at https://github.com/Fjr9516/GSL.
ABSTRACT
This paper is concentrated on the problem of adaptive fuzzy tracking control for an uncertain nonlinear system whose actuator is encountered by the asymmetric backlash behavior. First, we propose a new smooth inverse model which can approximate the asymmetric actuator backlash arbitrarily. By applying it, two adaptive fuzzy control scenarios, namely, the compensation-based control scheme and nonlinear decomposition-based control scheme, are then developed successively. It is worth noticing that the first fuzzy controller exhibits a better tracking control performance, although it recourses to a known slope ratio of backlash nonlinearity. The second one further removes the restriction, and also gets a desirable control performance. By the strict Lyapunov argument, both adaptive fuzzy controllers guarantee that the output tracking error is convergent to an adjustable region of zero asymptotically, while all the signals remain semiglobally uniformly ultimately bounded. Lastly, two comparative simulations are conducted to verify the effectiveness of the proposed fuzzy controllers.
ABSTRACT
To achieve better performance in multifocus image fusion problems, a new regional approach based on superpixels and superpixel-based mean filtering is proposed in this paper. First, a fast and effective segmentation method is adopted to generate the superpixels over a clarity-enhanced average image. By averaging the clarity information in each superpixel, we make the initial decision map of fusion by regionally selecting sharper superpixels in different source images. Then a novel superpixel-based mean filtering technique is introduced to make full use of spatial consistency in images and the final post-processed decision map is produced. The fused image is constructed by selecting pixels from different source images according to the final decision map. Experimental results demonstrate the proposed method's competitive performance in comparison with state-of-the-art multifocus image fusion approaches.
ABSTRACT
Although previous bio-inspired models have concentrated on invertebrates (such as ants), mammals such as primates with higher cognitive function are valuable for modeling the increasingly complex problems in engineering. Understanding primates' social and communication systems, and applying what is learned from them to engineering domains is likely to inspire solutions to a number of problems. This paper presents a novel bio-inspired approach to determine group size by researching and simulating primate society. Group size does matter for both primate society and digital entities. It is difficult to determine how to group mobile sensors/robots that patrol in a large area when many factors are considered such as patrol efficiency, wireless interference, coverage, inter/intragroup communications, etc. This paper presents a simulation-based theoretical study on patrolling strategies for robot groups with the comparison of large and small groups through simulations and theoretical results.