DSKCA-UNet: Dynamic selective kernel channel attention for medical image segmentation.

U-Net has attained immense popularity owing to its performance in medical image segmentation. However, it cannot be modeled explicitly over remote dependencies. By contrast, the transformer can effectively capture remote dependencies by leveraging the self-attention (SA) of the encoder. Although SA, an important characteristic of the transformer, can find correlations between them based on the original data, secondary computational complexity might retard the processing rate of high-dimensional data (such as medical images). Furthermore, SA is limited because the correlation between samples is overlooked; thus, there is considerable scope for improvement. To this end, based on Swin-UNet, we introduce a dynamic selective attention mechanism for the convolution kernels. The weight of each convolution kernel is calculated to fuse the results dynamically. This attention mechanism permits each neuron to adaptively modify its receptive field size in response to multiscale input information. A local cross-channel interaction strategy without dimensionality reduction was introduced, which effectively eliminated the influence of downscaling on learning channel attention. Through suitable cross-channel interactions, model complexity can be significantly reduced while maintaining its performance. Subsequently, the global interaction between the encoder features is used to extract more fine-grained features. Simultaneously, the mixed loss function of the weighted cross-entropy loss and Dice loss is used to alleviate category imbalances and achieve better results when the sample number is unbalanced. We evaluated our proposed method on abdominal multiorgan segmentation and cardiac segmentation datasets, achieving Dice similarity coefficient and 95% Hausdorff distance metrics of 80.30 and 14.55%, respectively, on the Synapse dataset and Dice similarity coefficient metrics of 90.80 on the ACDC dataset. The experimental results show that our proposed method has good generalization ability and robustness, and it is a powerful tool for medical image segmentation.

Feature interaction network based on hierarchical decoupled convolution for 3D medical image segmentation.

Manual image segmentation consumes time. An automatic and accurate method to segment multimodal brain tumors using context information rich three-dimensional medical images that can be used for clinical treatment decisions and surgical planning is required. However, it is a challenge to use deep learning to achieve accurate segmentation of medical images due to the diversity of tumors and the complex boundary interactions between sub-regions while limited computing resources hinder the construction of efficient neural networks. We propose a feature fusion module based on a hierarchical decoupling convolution network and an attention mechanism to improve the performance of network segmentation. We replaced the skip connections of U-shaped networks with a feature fusion module to solve the category imbalance problem, thus contributing to the segmentation of more complicated medical images. We introduced a global attention mechanism to further integrate the features learned by the encoder and explore the context information. The proposed method was evaluated for enhance tumor, whole tumor, and tumor core, achieving Dice similarity coefficient metrics of 0.775, 0.900, and 0.827, respectively, on the BraTS 2019 dataset and 0.800, 0.902, and 0.841, respectively on the BraTS 2018 dataset. The results show that our proposed method is inherently general and is a powerful tool for brain tumor image studies. Our code is available at: https://github.com/WSake/Feature-interaction-network-based-on-Hierarchical-Decoupled-Convolution.

Hypoxia and pH co-triggered oxidative stress amplifier for tumor therapy.

To keep fast proliferation, tumor cells are exposed to higher oxidative stress than normal cells and they upregulate the amount of some antioxidants such as glutathione (GSH) against reactive oxygen species to maintain the balance. This phenomenon is severe in hypoxic tumor cells. Although researchers have proposed a series of treatment strategies based on regulating the intracellular reactive oxygen species level, few of them are related to the hypoxic tumor. Herein, a novel organic compound (PLC) was designed by using lysine as a bridge to connect two functional small molecules, a hypoxia-responsive nitroimidazole derivative (pimonidazole) and a pH-responsive cinnamaldehyde (CA) derivative. Then, the oxidative stress amplifying ability of PLC in hypoxic tumor cells was evaluated. The acidic microenvironment of tumor can trigger the release of CA to produce reactive oxygen species. Meanwhile, large amount of nicotinamide adenine dinucleotide phosphate (NADPH) can be consumed to decrease the synthesis of GSH during the bio-reduction process of the nitro group in PLC under hypoxic conditions. Therefore, the lethal effect of CA can be amplified for the decrease of GSH. Our results prove that this strategy can significantly enhance the therapeutic effect of CA in the hypoxic tumor cells.

Hypoxia-Triggered Self-Assembly of Ultrasmall Iron Oxide Nanoparticles to Amplify the Imaging Signal of a Tumor.

Hypoxia is a common phenomenon among most solid tumors that significantly influences tumor response toward chemo- and radiotherapy. Understanding the distribution and extent of tumor hypoxia in patients will be very important to provide personalized therapies in the clinic. Without sufficient vessels, however, traditional contrast agents for clinical imaging techniques will have difficulty in accumulating in the hypoxic region of solid tumors, thus challenging the detection of hypoxia in vivo. To overcome this problem, herein we develop a novel hypoxia imaging probe, consisting of a hypoxia-triggered self-assembling ultrasmall iron oxide (UIO) nanoparticle and assembly-responding fluorescence dyes (NBD), to provide dual-mode imaging in vivo. In this strategy, we have employed nitroimidazole derivatives as the hypoxia-sensitive moiety to construct intermolecular cross-linking of UIO nanoparticles under hypoxia, which irreversibly form larger nanoparticle assemblies. The hypoxia-triggered performance of UIO self-assembly not only amplifies its T2-weighted MRI signal but also promotes the fluorescence intensity of NBD through its emerging hydrophobic environment incorporated into self-assemblies. In vivo results further confirm that our hypoxic imaging probe can display a prompt MRI signal for the tumor interior region, and its signal enhancement performs a long-term effective feature and gradually reaches 3.69 times amplification. Simultaneously, this probe also exhibits obvious green fluorescence in the hypoxic region of tumor sections. Accordingly, we also have developed a MRI difference value method to visualize the 3D distribution and describe the extent of the hypoxic tumor region within the whole bodies of mice. Due to its notable efficiency of penetration and accumulation inside a hypoxic tumor, our hypoxia imaging probe could also be considered as a potential candidate as a versatile platform for hypoxia-targeted drug delivery, and meanwhile its hypoxia-related therapeutic efficacy can be monitored.

Designing Hypoxia-Responsive Nanotheranostic Agents for Tumor Imaging and Therapy.

Hypoxia, a common feature of most solid tumors, plays an important role in tumor proliferation, metastasis, and invasion, leading to drug, radiation, and photodynamic therapy resistance, and resulting in a sharp reduction in the disease-free survival rate of tumor patients. The lack of sufficient blood supply to the interior regions of tumors hinders the delivery of traditional drugs and contrast agents, interfering with their accumulation in the hypoxic region, and preventing efficient theranostics. Thus, there is a need for the fabrication of novel tumor theranostic agents that overcome these obstacles. Reports, in recent years, of hypoxia-responsive nanomaterials may provide with such means. In this review, a comprehensive description of the physicochemical and biological characteristics of hypoxic tumor tissues is provided, the principles of designing the hypoxia-responsive tumor theranostic agents are discussed, and the recent research into hypoxia-triggered nanomaterials is examined. Additionally, other hypoxia-associated responsive strategies, the current limitations, and future prospects for hypoxia-responsive nanotheranostic agents in tumor treatment are discussed.