SCIENTIFIC COMPUTING AND IMAGING INSTITUTE
at the University of Utah

Advanced manufacturing creates increasingly complex objects with material compositions that are often difficult to characterize by a single modality. Our collaborating domain scientists are going beyond traditional methods by employing both X-ray and neutron computed tomography to obtain complementary representations expected to better resolve material boundaries. However, the use of two modalities creates its own challenges for visualization, requiring either complex adjustments of bimodal transfer functions or the need for multiple views. Together with experts in nondestructive evaluation, we designed a novel interactive bimodal visualization approach to create a combined view of the co-registered X-ray and neutron acquisitions of industrial objects. Using an automatic topological segmentation of the bivariate histogram of X-ray and neutron values as a starting point, the system provides a simple yet effective interface to easily create, explore, and adjust a bimodal visualization. We propose a widget with simple brushing interactions that enables the user to quickly correct the segmented histogram results. Our semiautomated system enables domain experts to intuitively explore large bimodal datasets without the need for either advanced segmentation algorithms or knowledge of visualization techniques. We demonstrate our approach using synthetic examples, industrial phantom objects created to stress bimodal scanning techniques, and real-world objects, and we discuss expert feedback.

Separating algorithms from their computation schedule has become a de facto solution to tackle the challenges of developing high performance code on modern heterogeneous architectures. Common approaches include Domain-specific languages (DSLs) which provide familiar APIs to domain experts, code generation frameworks that automate the generation of fast and portable code, and runtime systems that manage threads for concurrency and parallelism. In this paper, we present the Halide code generation framework for Phylanx distributed array processing platform. This extension enables compile-time optimization of Phylanx primitives for target architectures. To accomplish this, (1) we implemented new Phylanx primitives using Halide, and (2) partially exported Halide’s thread pool API to carry out parallelism on HPX (Phylanx’s runtime) threads. (3) showcased HPX performance analysis tools made available to Halide applications. The evaluation of the work has been done in two steps. First, we compare the performance of Halide applications running on its native runtime with that of the new HPX backend to verify there is no cost associated with using HPX threads. Next, we compare performances of a number of original implementations of Phylanx primitives against the new ones in Halide to verify performance and portability benefits of Halide in the context of Phylanx.

Statistical shape modeling (SSM) is a powerful computational framework for quantifying and analyzing the geometric variability of anatomical structures, facilitating advancements in medical research, diagnostics, and treatment planning. Traditional methods for shape modeling from imaging data demand significant manual and computational resources. Additionally, these methods necessitate repeating the entire modeling pipeline to derive shape descriptors (e.g., surface-based point correspondences) for new data. While deep learning approaches have shown promise in streamlining the construction of SSMs on new data, they still rely on traditional techniques to supervise the training of the deep networks. Moreover, the predominant linearity assumption of traditional approaches restricts their efficacy, a limitation also inherited by deep learning models trained using optimized/established correspondences. Consequently, representing complex anatomies becomes challenging. To address these limitations, we introduce SCorP, a novel framework capable of predicting surface-based correspondences directly from unsegmented images. By leveraging the shape prior learned directly from surface meshes in an unsupervised manner, the proposed model eliminates the need for an optimized shape model for training supervision. The strong shape prior acts as a teacher and regularizes the feature learning of the student network to guide it in learning image-based features that are predictive of surface correspondences. The proposed model streamlines the training and inference phases by removing the supervision for the correspondence prediction task while alleviating the linearity assumption. Experiments on the LGE MRI left atrium dataset and Abdomen CT-1K liver datasets demonstrate that the proposed technique enhances the accuracy and robustness of image-driven SSM, providing a compelling alternative to current fully supervised methods.

The study of physiology demonstrates that the form (shape) of anatomical structures dictates their functions, and analyzing the form of anatomies plays a crucial role in clinical research. Statistical shape modeling (SSM) is a widely used tool for quantitative analysis of forms of anatomies, aiding in characterizing and identifying differences within a population of subjects. Despite its utility, the conventional SSM construction pipeline is often complex and time-consuming. Additionally, reliance on linearity assumptions further limits the model from capturing clinically relevant variations. Recent advancements in deep learning solutions enable the direct inference of SSM from unsegmented medical images, streamlining the process and improving accessibility. However, the new methods of SSM from images do not adequately account for situations where the imaging data quality is poor or where only sparse information is available. Moreover, quantifying aleatoric uncertainty, which represents inherent data variability, is crucial in deploying deep learning for clinical tasks to ensure reliable model predictions and robust decision-making, especially in challenging imaging conditions. Therefore, we propose SPI-CorrNet, a unified model that predicts 3D correspondences from sparse imaging data. It leverages a teacher network to regularize feature learning and quantifies data-dependent aleatoric uncertainty by adapting the network to predict intrinsic input variances. Experiments on the LGE MRI left atrium dataset and Abdomen CT-1K liver datasets demonstrate that our technique enhances the accuracy and robustness of sparse image-driven SSM.

In the field of Nuclear Forensics, there exists a plethora of different tools to aid investigators when performing analysis of unknown nuclear materials. Many of these tools offer visual representations of the uranium ore concentrate (UOC) materials that include complimentary and contrasting information. In this paper, we present a novel technique drawing from state-of-the-art machine learning methods that allows information from scanning electron microscopy images (SEM) to be combined to create digital encodings of the material that can be used to determine the material’s processing route. Our technique can classify UOC processing routes with greater than 96% accuracy in a fraction of a second and can be adapted to unseen samples at similarly high accuracy. The technique’s high accuracy and speed allow forensic investigators to quickly get preliminary results, while generalization allows the model to be adapted to new materials or processing routes quickly without the need for complete retraining of the model.

The platelet integrin α_IIbβ₃ undergoes long-range conformational transitions between bent and extended conformations to regulate platelet aggregation during hemostasis and thrombosis. However, how exactly α_IIbβ₃ transitions between conformations remains largely elusive. Here, we studied how transitions across bent and extended-closed conformations of α_IIbβ₃ integrin are regulated by effective interactions between its functional domains. We first carried out μs-long equilibrium molecular dynamics (MD) simulations of full-length α_IIbβ₃ integrins in bent and intermediate conformations, the latter characterized by an extended headpiece and closed legs. Then, we built heterogeneous elastic network models, perturbed inter-domain interactions, and evaluated their relative contributions to the energy barriers between conformations. Results showed that integrin extension emerges from: (i) changes in interfaces between functional domains; (ii) allosteric coupling of the head and upper leg domains with flexible lower leg domains. Collectively, these results provide new insights into integrin conformational activation based on short- and long-range interactions between its functional domains and highlight the importance of the lower legs in the regulation of integrin allostery.

Hematoxylin and Eosin (H&E) staining is the most commonly used for disease diagnosis and tumor recurrence tracking. Hematoxylin excels at highlighting nuclei, whereas eosin stains the cytoplasm. However, H&E stain lacks details for differentiating different types of cells relevant to identifying the grade of the disease or response to specific treatment variations. Pathologists require special immunohistochemical (IHC) stains that highlight different cell types. These stains help in accurately identifying different regions of disease growth and their interactions with the cell’s microenvironment. The advent of deep learning models has made Image-to-Image (I2I) translation a key research area, reducing the need for expensive physical staining processes. Pix2Pix and CycleGAN are still the most commonly used methods for virtual staining applications. However, both suffer from hallucinations or staining irregularities when H&E stain has less discriminate information about the underlying cells IHC needs to highlight (e.g.,CD3 lymphocytes). Diffusion models are currently the state-of-the-art models for image generation and conditional generation tasks. However, they require extensive and diverse datasets (millions of samples) to converge, which is less feasible for virtual staining applications. Inspired by the success of multitask deep learning models for limited dataset size, we propose StainDiffuser, a novel multitask dual diffusion architecture for virtual staining that converges under a limited training budget. StainDiffuser trains two diffusion processes simultaneously: (a) generation of cell-specific IHC stain from H&E and (b) H&E-based cell segmentation using coarse segmentation only during training. Our results show that StainDiffuser produces high-quality results for easier (CK8/18,epithelial marker) and difficult stains(CD3, Lymphocytes).

How do cancer cells grow, divide, proliferate and die? How do drugs influence these processes? These are difficult questions that we can attempt to answer with a combination of time-series microscopy experiments, classification algorithms, and data visualization. However, collecting this type of data and applying algorithms to segment and track cells and construct lineages of proliferation is error-prone; and identifying the errors can be challenging since it often requires cross-checking multiple data types. Similarly, analyzing and communicating the results necessitates synthesizing different data types into a single narrative. State-of-the-art visualization methods for such data use independent line charts, tree diagrams, and images in separate views. However, this spatial separation requires the viewer of these charts to combine the relevant pieces of data in memory. To simplify this challenging task, we describe design principles for weaving cell images, time-series data, and tree data into a cohesive visualization. Our design principles are based on choosing a primary data type that drives the layout and integrates the other data types into that layout. We then introduce Aardvark, a system that uses these principles to implement novel visualization techniques. Based on Aardvark, we demonstrate the utility of each of these approaches for discovery, communication, and data debugging in a series of case studies.

Wildfires pose substantial risks to our health, environment, and economy. Studying wildfires is challenging due to their complex interaction with the atmosphere dynamics and the terrain. Researchers have employed ensemble simulations to study the relationship among variables and mitigate uncertainties in unpredictable initial conditions. However, many wildfire researchers are unaware of the advanced visualization available for conveying uncertainty. We designed and implemented an interactive visualization system for studying the uncertainties of fire spread patterns utilizing band-depth-based order statistics and contour boxplots. We also augment the visualization system with the summary of changes in the burned area and fuel content to help scientists identify interesting temporal events. In this paper, we demonstrate how our system can support wildfire experts in studying fire spread patterns, identifying outlier simulations, and navigating to interesting times based on a summary of events.

The recent introduction of multimodal large language models (MLLMs) combine the inherent power of large language models (LLMs) with the renewed capabilities to reason about the multimodal context. The potential usage scenarios for MLLMs significantly outpace their text-only counterparts. Many recent works in visualization have demonstrated MLLMs' capability to understand and interpret visualization results and explain the content of the visualization to users in natural language. In the machine learning community, the general vision capabilities of MLLMs have been evaluated and tested through various visual understanding benchmarks. However, the ability of MLLMs to accomplish specific visualization tasks based on visual perception has not been properly explored and evaluated, particularly, from a visualization-centric perspective.

In this work, we aim to fill the gap by utilizing the concept of visualization literacy to evaluate MLLMs. We assess MLLMs' performance over two popular visualization literacy evaluation datasets (VLAT and mini-VLAT). Under the framework of visualization literacy, we develop a general setup to compare different multimodal large language models (e.g., GPT4-o, Claude 3 Opus, Gemini 1.5 Pro) as well as against existing human baselines. Our study demonstrates MLLMs' competitive performance in visualization literacy, where they outperform humans in certain tasks such as identifying correlations, clusters, and hierarchical structures.

Idling vehicle detection (IVD) can be helpful in monitoring and reducing unnecessary idling and can be integrated into real-time systems to address the resulting pollution and harmful products. The previous approach, a non-end-to-end model, requires extra user clicks to specify a part of the input, making system deployment more error-prone or even not feasible. In contrast, we introduce an end-to-end joint audio-visual IVD task designed to detect vehicles visually under three states: moving, idling and engine off. Unlike feature co-occurrence task such as audio-visual vehicle tracking, our IVD task addresses complementary features, where labels cannot be determined by a single modality alone. To this end, we propose AVIVD-Net, a novel network that integrates audio and visual features through a bidirectional attention mechanism. AVIVD-Net streamlines the input process by learning a joint feature space, reducing the deployment complexity of previous methods. Additionally, we introduce the AVIVD dataset, which is seven times larger than previous datasets, offering significantly more annotated samples to study the IVD problem. Our model achieves performance comparable to prior approaches, making it suitable for automated deployment. Furthermore, by evaluating AVIVDNet on the feature co-occurrence public dataset MAVD, we demonstrate its potential for extension to self-driving vehicle video-camera setups.

The growing popularity of interactive time series exploration platforms has made visualizing data of public interest more accessible to general audiences. At the same time, the democratized access to professional-looking explorers with preloaded data enables the creation of convincing visualizations with carefully cherry-picked items. Prior research shows that people use data explorers to create and share charts that support their potentially biased or misleading views on public health or economic policy and that such charts have, for example, contributed to the spread of COVID-19 misinformation. Interventions against misinformation have focused on post hoc approaches such as fact-checking or removing misleading content, which are known to be challenging to execute. In this work, we explore whether we can use visualization design to impede cherry-picking—one of the most common methods employed by deceptive charts created on data exploration platforms. We describe a design space of guardrails—interventions against cherry-picking in time series explorers. Using our design space, we create a prototype data explorer with four types of guardrails and conduct two crowd-sourced experiments. In the first experiment, we challenge participants to create cherry-picked charts. We then use these charts in a second experiment to evaluate the guardrails’ impact on the perception of cherry-picking. We find evidence that guardrails—particularly superimposing relevant primary data—are successful at encouraging skepticism in a subset of experimental conditions but come with limitations. Based on our findings, we propose recommendations for developing effective guardrails for visualizations.

This paper introduces the Kernel Neural Operator (KNO), a novel operator learning technique that uses deep kernel-based integral operators in conjunction with quadrature for function-space approximation of operators (maps from functions to functions). KNOs use parameterized, closed-form, finitely-smooth, and compactly-supported kernels with trainable sparsity parameters within the integral operators to significantly reduce the number of parameters that must be learned relative to existing neural operators. Moreover, the use of quadrature for numerical integration endows the KNO with geometric flexibility that enables operator learning on irregular geometries. Numerical results demonstrate that on existing benchmarks the training and test accuracy of KNOs is higher than popular operator learning techniques while using at least an order of magnitude fewer trainable parameters. KNOs thus represent a new paradigm of low-memory, geometrically-flexible, deep operator learning, while retaining the implementation simplicity and transparency of traditional kernel methods from both scientific computing and machine learning.

Scalar field comparison is a fundamental task in scientific visualization. In topological data analysis, we compare topological descriptors of scalar fields—such as persistence diagrams and merge trees—because they provide succinct and robust abstract representations. Several similarity measures for topological descriptors seem to be both asymptotically and practically efficient with polynomial time algorithms, but they do not scale well when handling large-scale, time-varying scientific data and ensembles. In this paper, we propose a new framework to facilitate the comparative analysis of merge trees, inspired by tools from locality sensitive hashing (LSH). LSH hashes similar objects into the same hash buckets with high probability. We propose two new similarity measures for merge trees that can be computed via LSH, using new extensions to Recursive MinHash and subpath signature, respectively. Our similarity measures are extremely efficient to compute and closely resemble the results of existing measures such as merge tree edit distance or geometric interleaving distance. Our experiments demonstrate the utility of our LSH framework in applications such as shape matching, clustering, key event detection, and ensemble summarization.

Self-supervised learning has become a cornerstone in various areas, particularly histopathological image analysis. Image augmentation plays a crucial role in self-supervised learning, as it generates variations in image samples. However, traditional image augmentation techniques often overlook the unique characteristics of histopathological images. In this paper, we propose a new histopathology-specific image augmentation method called stain reconstruction augmentation (SRA). We integrate our SRA with MoCo v3, a leading model in self-supervised contrastive learning, along with our additional contrastive loss terms, and call the new model SRA-MoCo v3. We demonstrate that our SRA-MoCo v3 always outperforms the standard MoCo v3 across various downstream tasks and achieves comparable or superior performance to other foundation models pre-trained on significantly larger histopathology datasets.