Masked Autoencoder (MAE) is a self-supervised pre-training technique that holds promise in improving the representation learning of neural networks. However, the current application of MAE directly to volumetric medical images poses two challenges: (i) insufficient global information for clinical context understanding of the holistic data, and (ii) the absence of any assurance of stabilizing the representations learned from randomly masked inputs. To conquer these limitations, we propose the Global-Local Masked AutoEncoders (GL-MAE), a simple yet effective self-supervised pre-training strategy. GL-MAE acquires robust anatomical structure features by incorporating multi-level reconstruction from fine-grained local details to high-level global semantics. Furthermore, a complete global view serves as an anchor to direct anatomical semantic alignment and stabilize the learning process through global-to-global consistency learning and global-to-local consistency learning. Our fine-tuning results on eight mainstream public datasets demonstrate the superiority of our method over other state-of-the-art self-supervised algorithms, highlighting its effectiveness on versatile volumetric medical image segmentation and classification tasks. We will release codes upon acceptance at https://github.com/JiaxinZhuang/GL-MAE.

Many existing approaches for mammogram analysis are based on single view. Some recent DNN-based multi-view approaches can perform either bilateral or ipsilateral analysis, while in practice, radiologists use both to achieve the best clinical outcome. MommiNet is the first DNN-based tri-view mass identification approach, which can simultaneously perform bilateral and ipsilateral analysis of mammographic images, and in turn, can fully emulate the radiologists' reading practice. In this paper, we present MommiNet-v2, with improved network architecture and performance. Novel high-resolution network (HRNet)-based architectures are proposed to learn the symmetry and geometry constraints, to fully aggregate the information from all views for accurate mass detection. A multi-task learning scheme is adopted to incorporate both Breast Imaging-Reporting and Data System (BI-RADS) and biopsy information to train a mass malignancy classification network. Extensive experiments have been conducted on the public DDSM (Digital Database for Screening Mammography) dataset and our in-house dataset, and state-of-the-art results have been achieved in terms of mass detection accuracy. Satisfactory mass malignancy classification result has also been obtained on our in-house dataset.

Mammography plays an essential role in early detection of breast cancer. Interpreting mammography is a professional task that requires well-trained radiologists with longtime clinical experience. In this paper, we present MABEL, an artificial intelligence-powered system to assist doctors for breast cancer screening and diagnosis in mammograms, in order to reduce their workloads and accelerate the diagnostic process. Our system smoothly integrates our upgraded lesion identification models, provides a doctor-oriented annotation tool and web interface, and can communicate with Picture Archiving and Communication System (PACS) in our collaborative hospital. Our lesion identification performance is evaluated on both public and in-house datasets, in which mass detection has achieved state-of-the-art accuracy in the single-view manner. The overall high satisfaction from doctors of our system is also demonstrated.

In order to properly control slip-to-slip shift process of electric vehicles (EVs) equipped with dry dual-clutch transmissions (DDCT), dynamic modelling of a dry-type two-speed dual-clutch transmission for an electric vehicle in torque phase is proposed at first. Then, an optimal control strategy is proposed to investigate feasibility of non-shock shift process of torque phase without power interruption and power circulation. Optimal solutions of electric motor torque and dual clutch friction torques are derived in analytical form. And then, a dynamic model of fork-lever actuator is integrated into DDCT driveline dynamic model of EV, and an affine nonlinear shift dynamic model for the whole DDCT system is proposed to describe dynamic behaviours in torque phases of shift. Further, to solve problems in tracking inaccuracy induced by strong nonlinearities and modelling uncertainties ofthe dynamic model, sliding mode control strategy based on feedback linearization control theory is proposed. Accurate electric motor torque as well as motor currents imposed to affine nonlinear system are calculated through nonlinear feedback control law. Finally, tracking control accuracy of the affine nonlinear dynamic system is investigated through numerical simulation on MATLAB/Simulink platform. The simulation results verify that not only non-shock shift process of torque phase without power interruption and power circulation is realized, but also torque phase time can be adjusted to an arbitrary value based on actual requirement. Besides, high-precision tracings to optimal angular velocities of electric motor and dual clutch are realized by accurately modulating motor control currents and electric motor torque.