Shadow detection in general photos is a nontrivial problem, due to the complexity of the real world. Though recent shadow detectors have already achieved remarkable performance on various benchmark data, their performance is still limited for general real-world situations. In this work, we collected shadow images for multiple scenarios and compiled a new dataset of 10,500 shadow images, each with labeled ground-truth mask, for supporting shadow detection in the complex world. Our dataset covers a rich variety of scene categories, with diverse shadow sizes, locations, contrasts, and types. Further, we comprehensively analyze the complexity of the dataset, present a fast shadow detection network with a detail enhancement module to harvest shadow details, and demonstrate the effectiveness of our method to detect shadows in general situations.

Hybrid transformer-based segmentation approaches have shown great promise in medical image analysis. However, they typically require considerable computational power and resources during both training and inference stages, posing a challenge for resource-limited medical applications common in the field. To address this issue, we present an innovative framework called Slim UNETR, designed to achieve a balance between accuracy and efficiency by leveraging the advantages of both convolutional neural networks and transformers. Our method features the Slim UNETR Block as a core component, which effectively enables information exchange through self-attention mechanism decomposition and cost-effective representation aggregation. Additionally, we utilize the throughput metric as an efficiency indicator to provide feedback on model resource consumption. Our experiments demonstrate that Slim UNETR outperforms state-of-the-art models in terms of accuracy, model size, and efficiency when deployed on resource-constrained devices. Remarkably, Slim UNETR achieves 92.44% dice accuracy on BraTS2021 while being 34.4x smaller and 13.4x faster during inference compared to Swin UNETR. Code: https://github.com/aigzhusmart/Slim-UNETR.

Brain tumor segmentation from medical images is a prerequisite to provide a quantitative and intuitive reference for clinical diagnosis and treatment. Manual segmentation depends on clinicians’ experience, and is laborious and time-consuming. To tackle these issues, we proposed an encoder-decoder neural network, i.e. deep supervised 3D Squeeze-and-Excitation V-Net (DSSE-V-Net) to segment brain tumors automatically. We modified V-Net by adding batch normalization and using bottom residual block to make the network deeper. Then we incorporated a squeeze & excitation(SE) module in the modified V-Net by adding the SE block in each stage of the encoder and decoder, respectively. We also integrated 3D deep supervision seamlessly into the network to accelerate convergence. We evaluated our model on the public BraTS 2017 dataset for brain tumor segmentation. Our model outperformed both 3D U-Net and modified V-Net, and obtained highly competitive performance compared with those methods winning in the BraTS 2017 challenge.

Due to the individual difference, EEG signals from other subjects (source) can hardly be used to decode the mental intentions of the target subject. Although transfer learning methods have shown promising results, they still suffer from poor feature representation or neglect long-range dependencies. In light of these limitations, we propose Global Adaptive Transformer (GAT), an domain adaptation method to utilize source data for cross-subject enhancement. Our method uses parallel convolution to capture temporal and spatial features first. Then, we employ a novel attention-based adaptor that implicitly transfers source features to the target domain, emphasizing the global correlation of EEG features. We also use a discriminator to explicitly drive the reduction of marginal distribution discrepancy by learning against the feature extractor and the adaptor. Besides, an adaptive center loss is designed to align the conditional distribution. With the aligned source and target features, a classifier can be optimized to decode EEG signals. Experiments on two widely used EEG datasets demonstrate that our method outperforms state-of-the-art methods, primarily due to the effectiveness of the adaptor. These results indicate that GAT has good potential to enhance the practicality of BCI.

Ultrasound examination is widely used for diagnosis of breast cancer, which requires a full-coverage scan of the whole organ and stable probe–breast interaction for high-quality image acquisition. In this study, a robotic system for automated breast ultrasound scanning is proposed. To avoid occlusion, point cloud of the breast is obtained from multiple angles and registered together for accurate tissue shape reconstruction, then scan path planning is performed with isometric 3-D point cloud searching algorithm for full and uniform coverage. To reduce the adverse influence of large tissue deformation during ultrasound scanning, a contact force–strain regression model is built for tissue deformation estimation and used as a feedforward to correct the planned scanning path, and a probe–tissue interaction model with scanning resistance is built for normal probe orientation adjustment. To maintain normal probe orientation and stable contact force during robotic ultrasound scanning, a hybrid force–velocity control framework is proposed based on admittance control algorithm. Experiment results show that the robotic system can successfully maintain a constant contact force and keep the probe's posture along the normal orientation of probe–breast interaction both in in-plane direction and out-of-plane direction.

The Vision Transformer (ViT) has demonstrated remarkable performance in Self-Supervised Learning (SSL) for 3D medical image analysis. Masked AutoEncoder (MAE) for feature pre-training can further unleash the potential of ViT on various medical vision tasks. However, due to large spatial sizes with much higher dimensions of 3D medical images, the lack of hierarchical design for MAE may hinder the performance of downstream tasks. In this paper, we propose a novel Mask in Mask (MiM) pre-training framework for 3D medical images, which aims to advance MAE by learning discriminative representation from hierarchical visual tokens across varying scales. We introduce multiple levels of granularity for masked inputs from the volume, which are then reconstructed simultaneously ranging at both fine and coarse levels. Additionally, a cross-level alignment mechanism is applied to adjacent level volumes to enforce anatomical similarity hierarchically. Furthermore, we adopt a hybrid backbone to enhance the hierarchical representation learning efficiently during the pre-training. MiM was pre-trained on a large scale of available 3D volumetric images, i.e., Computed Tomography (CT) images containing various body parts. Extensive experiments on twelve public datasets demonstrate the superiority of MiM over other SSL methods in organ/tumor segmentation and disease classification. We further scale up the MiM to large pre-training datasets with more than 10k volumes, showing that large-scale pre-training can further enhance the performance of downstream tasks. Code is available at https://github.com/JiaxinZhuang/MiM.