Brain tissue and tumor extraction plays a pivotal role in medical care and clinical research. Leveraging the diverse and complementary information provided by different imaging modalities is crucial to the success of these applications. However, existing deep learning-based methods exhibit limitations in two major aspects. First, they ignore the incorporation of task-oriented regularization when fusing multi-modal features, leading to suboptimal latent space learning. Second, these methods tend to overlook the explicit modeling and exploitation of prediction uncertainty, despite the strong correlation between prediction uncertainty and errors. To address these issues, we propose a novel lightweight network with uncertainty-guided latent space refinement for multi-modal brain tissue and tumor extraction, called UMNet. Particularly, UMNet features a modality-specific uncertainty-regularized feature fusion module (M-SUM), which facilitates latent space refinement to enable a more informed and effective aggregation of complementary multi-modal information. Additionally, we design an uncertainty-enhanced loss function (U-Loss) to explicitly harness the connection between prediction uncertainty and errors. Experimental results demonstrate that our proposed UMNet achieves promising performance, outperforming state-of-the-art methods for both brain tissue and tumor extraction tasks.

Heterogeneous data captured by different scanning devices and imaging protocols can affect the generalization performance of the deep learning magnetic resonance (MR) reconstruction model. While a centralized training model is effective in mitigating this problem, it raises concerns about privacy protection. Federated learning is a distributed training paradigm that can utilize multi-institutional data for collaborative training without sharing data. However, existing federated learning MR image reconstruction methods rely on models designed manually by experts, which are complex and computationally expensive, suffering from performance degradation when facing heterogeneous data distributions. In addition, these methods give inadequate consideration to fairness issues, namely ensuring that the model's training does not introduce bias towards any specific dataset's distribution. To this end, this paper proposes a generalizable federated neural architecture search framework for accelerating MR imaging (GAutoMRI). Specifically, automatic neural architecture search is investigated for effective and efficient neural network representation learning of MR images from different centers. Furthermore, we design a fairness adjustment approach that can enable the model to learn features fairly from inconsistent distributions of different devices and centers, and thus facilitate the model to generalize well to the unseen center. Extensive experiments show that our proposed GAutoMRI has better performances and generalization ability compared with seven state-of-the-art federated learning methods. Moreover, the GAutoMRI model is significantly more lightweight, making it an efficient choice for MR image reconstruction tasks. The code will be made available at https://github.com/ternencewu123/GAutoMRI.

With the successful application of deep learning to magnetic resonance (MR) imaging, parallel imaging techniques based on neural networks have attracted wide attention. However, in the absence of high-quality, fully sampled datasets for training, the performance of these methods is limited. And the interpretability of models is not strong enough. To tackle this issue, this paper proposes a Physics-bAsed unsupeRvised Contrastive rEpresentation Learning (PARCEL) method to speed up parallel MR imaging. Specifically, PARCEL has a parallel framework to contrastively learn two branches of model-based unrolling networks from augmented undersampled multi-coil k-space data. A sophisticated co-training loss with three essential components has been designed to guide the two networks in capturing the inherent features and representations for MR images. And the final MR image is reconstructed with the trained contrastive networks. PARCEL was evaluated on two vivo datasets and compared to five state-of-the-art methods. The results show that PARCEL is able to learn essential representations for accurate MR reconstruction without relying on fully sampled datasets. The code will be made available at https://github.com/ternencewu123/PARCEL.

BACKGROUND: The expression changes of some proteins are associated with cancer progression, and can be used as biomarkers in cancer diagnosis. Automated systems have been frequently applied in the large-scale detection of protein biomarkers and have provided a valuable complement for wet-laboratory experiments. For example, our previous work used an immunohistochemical image-based machine learning classifier of protein subcellular locations to screen biomarker proteins that change locations in colon cancer tissues. The tool could recognize the location of biomarkers but did not consider the effect of protein expression level changes on the screening process. RESULTS: In this study, we built an automated classification model that recognizes protein expression levels in immunohistochemical images, and used the protein expression levels in combination with subcellular locations to screen cancer biomarkers. To minimize the effect of non-informative sections on the immunohistochemical images, we employed the representative image patches as input and applied a Wasserstein distance method to determine the number of patches. For the patches and the whole images, we compared the ability of color features, characteristic curve features, and deep convolutional neural network features to distinguish different levels of protein expression and employed deep learning and conventional classification models. Experimental results showed that the best classifier can achieve an accuracy of 73.72% and an F1-score of 0.6343. In the screening of protein biomarkers, the detection accuracy improved from 63.64 to 95.45% upon the incorporation of the protein expression changes. CONCLUSIONS: Machine learning can distinguish different protein expression levels and speed up their annotation in the future. Combining information on the expression patterns and subcellular locations of protein can improve the accuracy of automatic cancer biomarker screening. This work could be useful in discovering new cancer biomarkers for clinical diagnosis and research.

Breast density is widely adopted to reflect the likelihood of early breast cancer development. Existing methods of mammographic density classification either require steps of manual operations or achieve only moderate classification accuracy due to the limited model capacity. In this study, we present a radiomics approach based on dilated and attention-guided residual learning for the task of mammographic density classification. The proposed method was instantiated with two datasets, one clinical dataset and one publicly available dataset, and classification accuracies of 88.7 and 70.0 percent were obtained, respectively. Although the classification accuracy of the public dataset was lower than the clinical dataset, which was very likely related to the dataset size, our proposed model still achieved a better performance than the naive residual networks and several recently published deep learning-based approaches. Furthermore, we designed a multi-stream network architecture specifically targeting at analyzing the multi-view mammograms. Utilizing the clinical dataset, we validated that multi-view inputs were beneficial to the breast density classification task with an increase of at least 2.0 percent in accuracy and the different views lead to different model classification capacities. Our method has a great potential to be further developed and applied in computer-aided diagnosis systems. Our code is available at https://github.com/lich0031/Mammographic_Density_Classification.

Multi-contrast magnetic resonance (MR) image registration is useful in the clinic to achieve fast and accurate imaging-based disease diagnosis and treatment planning. Nevertheless, the efficiency and performance of the existing registration algorithms can still be improved. In this paper, we propose a novel unsupervised learning-based framework to achieve accurate and efficient multi-contrast MR image registration. Specifically, an end-to-end coarse-to-fine network architecture consisting of affine and deformable transformations is designed to improve the robustness and achieve end-to-end registration. Furthermore, a dual consistency constraint and a new prior knowledge-based loss function are developed to enhance the registration performances. The proposed method has been evaluated on a clinical dataset containing 555 cases, and encouraging performances have been achieved. Compared to the commonly utilized registration methods, including VoxelMorph, SyN, and LT-Net, the proposed method achieves better registration performance with a Dice score of 0.8397± 0.0756 in identifying stroke lesions. With regards to the registration speed, our method is about 10 times faster than the most competitive method of SyN (Affine) when testing on a CPU. Moreover, we prove that our method can still perform well on more challenging tasks with lacking scanning information data, showing the high robustness for the clinical application.