Cardiovascular disease (CVD) is the leading cause of death worldwide, and myocardial perfusion imaging using SPECT has been widely used in the diagnosis of CVDs. The GE 530/570c dedicated cardiac SPECT scanners adopt a stationary geometry to simultaneously acquire 19 projections to increase sensitivity and achieve dynamic imaging. However, the limited amount of angular sampling negatively affects image quality. Deep learning methods can be implemented to produce higher-quality images from stationary data. This is essentially a few-view imaging problem. In this work, we propose a novel 3D transformer-based dual-domain network, called TIP-Net, for high-quality 3D cardiac SPECT image reconstructions. Our method aims to first reconstruct 3D cardiac SPECT images directly from projection data without the iterative reconstruction process by proposing a customized projection-to-image domain transformer. Then, given its reconstruction output and the original few-view reconstruction, we further refine the reconstruction using an image-domain reconstruction network. Validated by cardiac catheterization images, diagnostic interpretations from nuclear cardiologists, and defect size quantified by an FDA 510(k)-cleared clinical software, our method produced images with higher cardiac defect contrast on human studies compared with previous baseline methods, potentially enabling high-quality defect visualization using stationary few-view dedicated cardiac SPECT scanners.

Sign language recognition (SLR) is a weakly supervised task that annotates sign videos as textual glosses. Recent studies show that insufficient training caused by the lack of large-scale available sign datasets becomes the main bottleneck for SLR. Most SLR works thereby adopt pretrained visual modules and develop two mainstream solutions. The multi-stream architectures extend multi-cue visual features, yielding the current SOTA performances but requiring complex designs and might introduce potential noise. Alternatively, the advanced single-cue SLR frameworks using explicit cross-modal alignment between visual and textual modalities are simple and effective, potentially competitive with the multi-cue framework. In this work, we propose a novel contrastive visual-textual transformation for SLR, CVT-SLR, to fully explore the pretrained knowledge of both the visual and language modalities. Based on the single-cue cross-modal alignment framework, we propose a variational autoencoder (VAE) for pretrained contextual knowledge while introducing the complete pretrained language module. The VAE implicitly aligns visual and textual modalities while benefiting from pretrained contextual knowledge as the traditional contextual module. Meanwhile, a contrastive cross-modal alignment algorithm is designed to explicitly enhance the consistency constraints. Extensive experiments on public datasets (PHOENIX-2014 and PHOENIX-2014T) demonstrate that our proposed CVT-SLR consistently outperforms existing single-cue methods and even outperforms SOTA multi-cue methods.

Large language models (LLM) not only empower multiple language tasks but also serve as a general interface across different spaces. Up to now, it has not been demonstrated yet how to effectively translate the successes of LLMs in the computer vision field to the medical imaging field which involves high-dimensional and multi-modal medical images. In this paper, we report a feasibility study of building a multi-task CT large image-text (LIT) model for lung cancer diagnosis by combining an LLM and a large image model (LIM). Specifically, the LLM and LIM are used as encoders to perceive multi-modal information under task-specific text prompts, which synergizes multi-source information and task-specific and patient-specific priors for optimized diagnostic performance. The key components of our LIT model and associated techniques are evaluated with an emphasis on 3D lung CT analysis. Our initial results show that the LIT model performs multiple medical tasks well, including lung segmentation, lung nodule detection, and lung cancer classification. Active efforts are in progress to develop large image-language models for superior medical imaging in diverse applications and optimal patient outcomes.

Recent research has revealed that Graph Neural Networks (GNNs) are susceptible to adversarial attacks targeting the graph structure. A malicious attacker can manipulate a limited number of edges, given the training labels, to impair the victim model's performance. Previous empirical studies indicate that gradient-based attackers tend to add edges rather than remove them. In this paper, we present a theoretical demonstration revealing that attackers tend to increase inter-class edges due to the message passing mechanism of GNNs, which explains some previous empirical observations. By connecting dissimilar nodes, attackers can more effectively corrupt node features, making such attacks more advantageous. However, we demonstrate that the inherent smoothness of GNN's message passing tends to blur node dissimilarity in the feature space, leading to the loss of crucial information during the forward process. To address this issue, we propose a novel surrogate model with multi-level propagation that preserves the node dissimilarity information. This model parallelizes the propagation of unaggregated raw features and multi-hop aggregated features, while introducing batch normalization to enhance the dissimilarity in node representations and counteract the smoothness resulting from topological aggregation. Our experiments show significant improvement with our approach.Furthermore, both theoretical and experimental evidence suggest that adding inter-class edges constitutes an easily observable attack pattern. We propose an innovative attack loss that balances attack effectiveness and imperceptibility, sacrificing some attack effectiveness to attain greater imperceptibility. We also provide experiments to validate the compromise performance achieved through this attack loss.

The large language model called ChatGPT has drawn extensively attention because of its human-like expression and reasoning abilities. In this study, we investigate the feasibility of using ChatGPT in experiments on using ChatGPT to translate radiology reports into plain language for patients and healthcare providers so that they are educated for improved healthcare. Radiology reports from 62 low-dose chest CT lung cancer screening scans and 76 brain MRI metastases screening scans were collected in the first half of February for this study. According to the evaluation by radiologists, ChatGPT can successfully translate radiology reports into plain language with an average score of 4.27 in the five-point system with 0.08 places of information missing and 0.07 places of misinformation. In terms of the suggestions provided by ChatGPT, they are general relevant such as keeping following-up with doctors and closely monitoring any symptoms, and for about 37% of 138 cases in total ChatGPT offers specific suggestions based on findings in the report. ChatGPT also presents some randomness in its responses with occasionally over-simplified or neglected information, which can be mitigated using a more detailed prompt. Furthermore, ChatGPT results are compared with a newly released large model GPT-4, showing that GPT-4 can significantly improve the quality of translated reports. Our results show that it is feasible to utilize large language models in clinical education, and further efforts are needed to address limitations and maximize their potential.

Deep learning (DL) has emerged as a new approach in the field of computed tomography (CT) with many applicaitons. A primary example is CT reconstruction from incomplete data, such as sparse-view image reconstruction. However, applying DL to sparse-view cone-beam CT (CBCT) remains challenging. Many models learn the mapping from sparse-view CT images to the ground truth but often fail to achieve satisfactory performance. Incorporating sinogram data and performing dual-domain reconstruction improve image quality with artifact suppression, but a straightforward 3D implementation requires storing an entire 3D sinogram in memory and many parameters of dual-domain networks. This remains a major challenge, limiting further research, development and applications. In this paper, we propose a sub-volume-based 3D denoising diffusion probabilistic model (DDPM) for CBCT image reconstruction from down-sampled data. Our DDPM network, trained on data cubes extracted from paired fully sampled sinograms and down-sampled sinograms, is employed to inpaint down-sampled sinograms. Our method divides the entire sinogram into overlapping cubes and processes them in parallel on multiple GPUs, successfully overcoming the memory limitation. Experimental results demonstrate that our approach effectively suppresses few-view artifacts while preserving textural details faithfully.

Since 2016, deep learning (DL) has advanced tomographic imaging with remarkable successes, especially in low-dose computed tomography (LDCT) imaging. Despite being driven by big data, the LDCT denoising and pure end-to-end reconstruction networks often suffer from the black box nature and major issues such as instabilities, which is a major barrier to apply deep learning methods in low-dose CT applications. An emerging trend is to integrate imaging physics and model into deep networks, enabling a hybridization of physics/model-based and data-driven elements. %This type of hybrid methods has become increasingly influential. In this paper, we systematically review the physics/model-based data-driven methods for LDCT, summarize the loss functions and training strategies, evaluate the performance of different methods, and discuss relevant issues and future directions.

Although contextualized embeddings generated from large-scale pre-trained models perform well in many tasks, traditional static embeddings (e.g., Skip-gram, Word2Vec) still play an important role in low-resource and lightweight settings due to their low computational cost, ease of deployment, and stability. In this paper, we aim to improve word embeddings by 1) incorporating more contextual information from existing pre-trained models into the Skip-gram framework, which we call Context-to-Vec; 2) proposing a post-processing retrofitting method for static embeddings independent of training by employing priori synonym knowledge and weighted vector distribution. Through extrinsic and intrinsic tasks, our methods are well proven to outperform the baselines by a large margin.

Pretrained protein structure models without labels are crucial foundations for the majority of protein downstream applications. The conventional structure pretraining methods follow the mature natural language pretraining methods such as denoised reconstruction and masked language modeling but usually destroy the real representation of spatial structures. The other common pretraining methods might predict a fixed set of predetermined object categories, where a restricted supervised manner limits their generality and usability as additional labeled data is required to specify any other protein concepts. In this work, we introduce a novel unsupervised protein structure representation pretraining with a robust protein language model. In particular, we first propose to leverage an existing pretrained language model to guide structure model learning through an unsupervised contrastive alignment. In addition, a self-supervised structure constraint is proposed to further learn the intrinsic information about the structures. With only light training data, the pretrained structure model can obtain better generalization ability. To quantitatively evaluate the proposed structure models, we design a series of rational evaluation methods, including internal tasks (e.g., contact map prediction, distribution alignment quality) and external/downstream tasks (e.g., protein design). The extensive experimental results conducted on multiple tasks and specific datasets demonstrate the superiority of the proposed sequence-structure transformation framework.

Artificial intelligence has made great progress in medical data analysis, but the lack of robustness and trustworthiness has kept these methods from being widely deployed. As it is not possible to train networks that are accurate in all situations, models must recognize situations where they cannot operate confidently. Bayesian deep learning methods sample the model parameter space to estimate uncertainty, but these parameters are often subject to the same vulnerabilities, which can be exploited by adversarial attacks. We propose a novel ensemble approach based on feature decorrelation and Fourier partitioning for teaching networks diverse complementary features, reducing the chance of perturbation-based fooling. We test our approach on electrocardiogram classification, demonstrating superior accuracy confidence measurement, on a variety of adversarial attacks. For example, on our ensemble trained with both decorrelation and Fourier partitioning scored a 50.18% inference accuracy and 48.01% uncertainty accuracy (area under the curve) on {\epsilon} = 50 projected gradient descent attacks, while a conventionally trained ensemble scored 21.1% and 30.31% on these metrics respectively. Our approach does not require expensive optimization with adversarial samples and can be scaled to large problems. These methods can easily be applied to other tasks for more robust and trustworthy models.