Wei-Ning Lee

Andong Wang, Wei-Ning Lee, Xiaojuan Qi

To interpret deep networks, one main approach is to associate neurons with human-understandable concepts. However, existing methods often ignore the inherent relationships of different concepts (e.g., dog and cat both belong to animals), and thus lose the chance to explain neurons responsible for higher-level concepts (e.g., animal). In this paper, we study hierarchical concepts inspired by the hierarchical cognition process of human beings. To this end, we propose HIerarchical Neuron concepT explainer (HINT) to effectively build bidirectional associations between neurons and hierarchical concepts in a low-cost and scalable manner. HINT enables us to systematically and quantitatively study whether and how the implicit hierarchical relationships of concepts are embedded into neurons, such as identifying collaborative neurons responsible to one concept and multimodal neurons for different concepts, at different semantic levels from concrete concepts (e.g., dog) to more abstract ones (e.g., animal). Finally, we verify the faithfulness of the associations using Weakly Supervised Object Localization, and demonstrate its applicability in various tasks such as discovering saliency regions and explaining adversarial attacks. Code is available on https://github.com/AntonotnaWang/HINT.

Andong Wang, Terence Chi Chun Tam, Ho Ming Poon et al.

Airway segmentation is essential for chest CT image analysis. Different from natural image segmentation, which pursues high pixel-wise accuracy, airway segmentation focuses on topology. The task is challenging not only because of its complex tree-like structure but also the severe pixel imbalance among airway branches of different generations. To tackle the problems, we present a NaviAirway method which consists of a bronchiole-sensitive loss function for airway topology preservation and an iterative training strategy for accurate model learning across different airway generations. To supplement the features of airway branches learned by the model, we distill the knowledge from numerous unlabeled chest CT images in a teacher-student manner. Experimental results show that NaviAirway outperforms existing methods, particularly in the identification of higher-generation bronchioles and robustness to new CT scans. Moreover, NaviAirway is general enough to be combined with different backbone models to significantly improve their performance. NaviAirway can generate an airway roadmap for Navigation Bronchoscopy and can also be applied to other scenarios when segmenting fine and long tubular structures in biomedical images. The code is publicly available on https://github.com/AntonotnaWang/NaviAirway.

Andong Wang, Wei-Ning Lee

To interpret deep learning models, one mainstream is to explore the learned concepts by networks. Testing with Concept Activation Vector (TCAV) presents a powerful tool to quantify the contribution of query concepts (represented by user-defined guidance images) to a target class. For example, we can quantitatively evaluate whether and to what extent concept striped contributes to model prediction zebra with TCAV. Therefore, TCAV whitens the reasoning process of deep networks. And it has been applied to solve practical problems such as diagnosis. However, for some images where the target object only occupies a small fraction of the region, TCAV evaluation may be interfered with by redundant background features because TCAV calculates concept contribution to a target class based on a whole hidden layer. To tackle this problem, based on TCAV, we propose Spatial Activation Concept Vector (SACV) which identifies the relevant spatial locations to the query concept while evaluating their contributions to the model prediction of the target class. Experiment shows that SACV generates a more fine-grained explanation map for a hidden layer and quantifies concepts' contributions spatially. Moreover, it avoids interference from background features. The code is available on https://github.com/AntonotnaWang/Spatial-Activation-Concept-Vector.

Jiajing Zhang, Bingze Dai, Xi Zhang et al.

The inherent electronic and speckle noise complicates clinical interpretation of ultrasound images. Conventional denoising methods rely on explicit noise assumptions whose validity diminishes under composite noise conditions. Learning-based methods require massive labeled data and model parameters. These pre-defined and pre-trained manners entail an inevitable domain shift in complex in vivo environments, so they are limited to a specific noise type and often blur structural details. In this study, we propose a pure test-time training framework for one-shot ultrasound image denoising and apply it to synthetic aperture ultrasound (SAU), which synthesizes transmit focus from sub-aperture transmissions. Our Aperture-to-Aperture (A2A) framework disentangles anatomical similarity and noise randomness from shuffled sub-apertures through self-contrastive learning in pyramid latent spaces. The clean image is then decoded from the anatomy space, while discarding the noise space. A2A is trained at test time on one noisy sample of SAU signals, so it fundamentally eliminates the domain shift and pretraining costs. Simulation experiments, including electronic noise levels of 0 to 30 dB and different inclusion geometries, demonstrated an improvement of 69.3% SNR and 34.4% CNR by A2A. The in vivo results showed 84.8% SNR and 25.7% CNR gains using only two aperture data of the heart in six echocardiographic views, liver, and kidney. A2A delivers clear images/signals across diverse imaging targets and configurations, paving the way for more reliable anatomical visualization and functional assessment by ultrasound.

Haotian Guan, Jinping Dong, Wei-Ning Lee

Ultrafast ultrasound blood flow imaging is a state-of-the-art technique for depiction of complex blood flow dynamics in vivo through thousands of full-view image data (or, timestamps) acquired per second. Physics-informed Neural Network (PINN) is one of the most preeminent solvers of the Navier-Stokes equations, widely used as the governing equation of blood flow. However, that current approaches rely on full Navier-Stokes equations is impractical for ultrafast ultrasound. We hereby propose a novel PINN training framework for solving the Navier-Stokes equations. It involves discretizing Navier-Stokes equations into steady state and sequentially solving them with test-time adaptation. The novel training framework is coined as SeqPINN. Upon its success, we propose a parallel training scheme for all timestamps based on averaged constant stochastic gradient descent as initialization. Uncertainty estimation through Stochastic Weight Averaging Gaussian is then used as an indicator of generalizability of the initialization. This algorithm, named SP-PINN, further expedites training of PINN while achieving comparable accuracy with SeqPINN. The performance of SeqPINN and SP-PINN was evaluated through finite-element simulations and in vitro phantoms of single-branch and trifurcate blood vessels. Results show that both algorithms were manyfold faster than the original design of PINN, while respectively achieving Root Mean Square Errors of 0.63 cm/s and 0.81 cm/s on the straight vessel and 1.35 cm/s and 1.63 cm/s on the trifurcate vessel when recovering blood flow velocities. The successful implementation of SeqPINN and SP-PINN open the gate for real-time training of PINN for Navier-Stokes equations and subsequently reliable imaging-based blood flow assessment in clinical practice.

Andong Wang, Bo Wu, Sunli Chen et al.

Learning commonsense reasoning from visual contexts and scenes in real-world is a crucial step toward advanced artificial intelligence. However, existing video reasoning benchmarks are still inadequate since they were mainly designed for factual or situated reasoning and rarely involve broader knowledge in the real world. Our work aims to delve deeper into reasoning evaluations, specifically within dynamic, open-world, and structured context knowledge. We propose a new benchmark (SOK-Bench), consisting of 44K questions and 10K situations with instance-level annotations depicted in the videos. The reasoning process is required to understand and apply situated knowledge and general knowledge for problem-solving. To create such a dataset, we propose an automatic and scalable generation method to generate question-answer pairs, knowledge graphs, and rationales by instructing the combinations of LLMs and MLLMs. Concretely, we first extract observable situated entities, relations, and processes from videos for situated knowledge and then extend to open-world knowledge beyond the visible content. The task generation is facilitated through multiple dialogues as iterations and subsequently corrected and refined by our designed self-promptings and demonstrations. With a corpus of both explicit situated facts and implicit commonsense, we generate associated question-answer pairs and reasoning processes, finally followed by manual reviews for quality assurance. We evaluated recent mainstream large vision-language models on the benchmark and found several insightful conclusions. For more information, please refer to our benchmark at www.bobbywu.com/SOKBench.

Xiaofei Sun, He Li, Wei-Ning Lee

Objective. A phased or a curvilinear array produces ultrasound (US) images with a sector field of view (FOV), which inherently exhibits spatially-varying image resolution with inferior quality in the far zone and towards the two sides azimuthally. Sector US images with improved spatial resolutions are favorable for accurate quantitative analysis of large and dynamic organs, such as the heart. Therefore, this study aims to translate US images with spatially-varying resolution to ones with less spatially-varying resolution. CycleGAN has been a prominent choice for unpaired medical image translation; however, it neither guarantees structural consistency nor preserves backscattering patterns between input and generated images for unpaired US images. Approach. To circumvent this limitation, we propose a constrained CycleGAN (CCycleGAN), which directly performs US image generation with unpaired images acquired by different ultrasound array probes. In addition to conventional adversarial and cycle-consistency losses of CycleGAN, CCycleGAN introduces an identical loss and a correlation coefficient loss based on intrinsic US backscattered signal properties to constrain structural consistency and backscattering patterns, respectively. Instead of post-processed B-mode images, CCycleGAN uses envelope data directly obtained from beamformed radio-frequency signals without any other non-linear postprocessing. Main Results. In vitro phantom results demonstrate that CCycleGAN successfully generates images with improved spatial resolution as well as higher peak signal-to-noise ratio (PSNR) and structural similarity (SSIM) compared with benchmarks. Significance. CCycleGAN-generated US images of the in vivo human beating heart further facilitate higher quality heart wall motion estimation than benchmarks-generated ones, particularly in deep regions.