Search

• Introducing three-dimension convolutional neural network can improve the performance of automated detection of focal cortical dysplasia. • Introducing positron emission tomography data aids in detecting lesions of focal cortical dysplasia. • Automated detection of focal cortical dysplasia with high sensitivity and few false-positive findings is feasible based on multi-modal data. Purpose: Focal cortical dysplasia (FCD) is a common etiology of drug-resistant focal epilepsy. Visual identification of FCD is usually time-consuming and depends on personal experience. Herein, we propose an automated type II FCD detection approach utilizing multi-modal data and 3D convolutional neural network (CNN). Methods: MRI and positron emission tomography (PET) data of 82 patients with FCD were collected, including 55 (67.1%) histopathologically, and 27 (32.9%) radiologically diagnosed patients. Three types of morphometric feature maps and three types of tissue maps were extracted from the T1-weighted images. These maps, T1, and PET images formed the inputs for CNN. Five-fold cross-validations were carried out on the training set containing 62 patients, and the model behaving best was chosen to detect FCD on the test set of 20 patients. Furthermore, ablation experiments were performed to estimate the value of PET data and CNN. Results: On the validation set, FCD was detected in 90.3% of the cases, with an average of 1.7 possible lesions per patient. The sensitivity on the test set was 90.0%, with 1.85 possible lesions per patient. Without the PET data, the sensitivity decreased to 80.0%, and the average lesion number increased to 2.05 on the test set. If an artificial neural network replaced the CNN, the sensitivity decreased to 85.0%, and the average lesion number increased to 4.65. Significance: Automated detection of FCD with high sensitivity and few false-positive findings is feasible based on multi-modal data. PET data and CNN could improve the performance of automated detection. [ABSTRACT FROM AUTHOR]

Abstract The tracking control problem of an underactuated ship is investigated. We intend to use the underactuated ship as an example to extend the traditional dynamic inversion control method to underactuated systems. The difficulty lies in the fact that the system has no relative degree, which prevents the application of standard dynamic inversion. Three modified dynamic inversion methods are proposed that are applicable to this system. The first is the well-known dynamic extension-based dynamic inversion (DEDI), which treats an input as a state and takes dynamic extension to achieve a relative degree. The second is virtual input-based dynamic inversion (VIDI), which treats a state as a virtual input to achieve a relative degree. The third is output redefinition-based dynamic inversion (ORDI), which selects a particular variable as a new output to achieve a relative degree. The three methods are generalizations of dynamic inversion control and remove some of its inherent limitations, making it applicable to a wide variety of underactuated systems. The effectiveness of the proposed methods is verified by numerical simulations. Highlights • The underactuated ship tracking control problem is solved in a clear and beautiful manner by using modified dynamic inversion. • The proposed method inherits the advantage of dynamic inversion control which does not require a careful construction of Lyapunov functions. • A unified framework of modified dynamic inversion is built, which removes some inherent limitations of dynamic inversion control, making it applicable to a wide variety of underactuated systems. [ABSTRACT FROM AUTHOR]

In this paper, the nonminimum phase problem of a flexible hypersonic vehicle is investigated. The main challenge of nonminimum phase is the prevention of dynamic inversion methods to nonlinear control design. To solve this problem, we make research on the relationship between nonminimum phase and backstepping control, finding that a stable nonlinear controller can be obtained by changing the control loop on the basis of backstepping control. By extending the control loop to cover the internal dynamics in it, the internal states are directly controlled by the inputs and simultaneously serve as virtual control for the external states, making it possible to guarantee output tracking as well as internal stability. Then, based on the extended control loop, a simplified control-oriented model is developed to enable the applicability of adaptive backstepping method. It simplifies the design process and releases some limitations caused by direct use of the no simplified control-oriented model. Next, under proper assumptions, asymptotic stability is proved for constant commands, while bounded stability is proved for varying commands. The proposed method is compared with approximate backstepping control and dynamic surface control and is shown to have superior tracking accuracy as well as robustness from the simulation results. This paper may also provide a beneficial guidance for control design of other complex systems. [ABSTRACT FROM AUTHOR]