Alark Joshi

Unified Framework for Development, Deployment and Robust Testing of Neuroimaging Algorithms

Developing both graphical and command-line user interfaces for neuroimaging algorithms requires considerable effort. Neuroimaging algorithms can meet their potential only if they can be easily and frequently used by their intended users. Deployment of a large suite of such algorithms on multiple platforms requires consistency of user interface controls, consistent results across various platforms and thorough testing. We present the design and implementation of a novel object-oriented framework that allows for rapid development of complex image analysis algorithms with many reusable components and the ability to easily add graphical user interface controls. Our framework also allows for simplified yet robust nightly testing of the algorithms to ensure stability and cross platform interoperability. All of the functionality is encapsulated into a software object requiring no separate source code for user interfaces, testing or deployment. This formulation makes our framework ideal for developing novel, stable and easy-to-use algorithms for medical image analysis and computer assisted interventions. The framework has been both deployed at Yale and released for public use in the open source multi-platform image analysis software—BioImage Suite (bioimagesuite.org).

Illustration-inspired techniques for visualizing time-varying data

Traditionally, time-varying data has been visualized using snapshots of the individual time steps or an animation of the snapshots shown in a sequential manner. For larger datasets with many time-varying features, animation can be limited in its use, as an observer can only track a limited number of features over the last few frames. Visually inspecting each snapshot is not practical either for a large number of time-steps. We propose new techniques inspired from the illustration literature to convey change over time more effectively in a time-varying dataset. Speedlines are used extensively by cartoonists to convey motion, speed, or change over different panels. Flow ribbons are another technique used by cartoonists to depict motion in a single frame. Strobe silhouettes are used to depict previous positions of an object to convey the previous positions of the object to the user. These illustration-inspired techniques can be used in conjunction with animation to convey change over time.

Texture-based feature tracking for effective time-varying data visualization

Analyzing, visualizing, and illustrating changes within time-varying volumetric data is challenging due to the dynamic changes occurring between timesteps. The changes and variations in computational fluid dynamic volumes and atmospheric 3D datasets do not follow any particular transformation. Features within the data move at different speeds and directions making the tracking and visualization of these features a difficult task. We introduce a texture-based feature tracking technique to overcome some of the current limitations found in the illustration and visualization of dynamic changes within time-varying volumetric data. Our texture-based technique tracks various features individually and then uses the tracked objects to better visualize structural changes. We show the effectiveness of our texture-based tracking technique with both synthetic and real world time-varying data. Furthermore, we highlight the specific visualization, annotation, registration, and feature isolation benefits of our technique. For instance, we show how our texture-based tracking can lead to insightful visualizations of time-varying data. Such visualizations, more than traditional visualization techniques, can assist domain scientists to explore and understand dynamic changes.

Rapid Development of Medical Imaging Tools with Open-Source Libraries

Rapid prototyping is an important element in researching new imaging analysis techniques and developing custom medical applications. In the last ten years, the open source community and the number of open source libraries and freely available frameworks for biomedical research have grown significantly. What they offer are now considered standards in medical image analysis, computer-aided diagnosis, and medical visualization. A cursory review of the peer-reviewed literature in imaging informatics (indeed, in almost any information technology-dependent scientific discipline) indicates the current reliance on open source libraries to accelerate development and validation of processes and techniques. In this survey paper, we review and compare a few of the most successful open source libraries and frameworks for medical application development. Our dual intentions are to provide evidence that these approaches already constitute a vital and essential part of medical image analysis, diagnosis, and visualization and to motivate the reader to use open source libraries and software for rapid prototyping of medical applications and tools.

Novel interaction techniques for neurosurgical planning and stereotactic navigation

Neurosurgical planning and image guided neurosurgery require the visualization of multimodal data obtained from various functional and structural image modalities, such as magnetic resonance imaging (MRI), computed tomography (CT), functional MRI, Single photon emission computed tomography (SPECT) and so on. In the case of epilepsy neurosurgery for example, these images are used to identify brain regions to guide intracranial electrode implantation and resection. Generally, such data is visualized using 2D slices and in some cases using a 3D volume rendering along with the functional imaging results. Visualizing the activation region effectively by still preserving sufficient surrounding brain regions for context is exceedingly important to neurologists and surgeons. We present novel interaction techniques for visualization of multimodal data to facilitate improved exploration and planning for neurosurgery. We extended the line widget from VTK to allow surgeons to control the shape of the region of the brain that they can visually crop away during exploration and surgery. We allow simple spherical, cubical, ellipsoidal and cylindrical (probe aligned cuts) for exploration purposes. In addition we integrate the cropping tool with the image-guided navigation system used for epilepsy neurosurgery. We are currently investigating the use of these new tools in surgical planning and based on further feedback from our neurosurgeons we will integrate them into the setup used for image-guided neurosurgery.

Case Study on Visualizing Hurricanes Using Illustration-Inspired Techniques

The devastating power of hurricanes was evident during the 2005 hurricane season, the most active season on record. This has prompted increased efforts by researchers to understand the physical processes that underlie the genesis, intensification, and tracks of hurricanes. This research aims at facilitating an improved understanding into the structure of hurricanes with the aid of visualization techniques. Our approach was developed by a mixed team of visualization and domain experts. To better understand these systems, and to explore their representation in NWP models, we use a variety of illustration-inspired techniques to visualize their structure and time evolution. Illustration-inspired techniques aid in the identification of the amount of vertical wind shear in a hurricane, which can help meteorologists predict dissipation. Illustration-style visualization, in combination with standard visualization techniques, helped explore the vortex rollup phenomena and the mesovortices contained within. We evaluated the effectiveness of our visualization with the help of six hurricane experts. The expert evaluation showed that the illustration-inspired techniques were preferred over existing tools. Visualization of the evolution of structural features is a prelude to a deeper visual analysis of the underlying dynamics.

Evaluation of illustration-inspired techniques for time-varying data visualization

Illustration‐inspired techniques have provided alternative ways to visualize time‐varying data. Techniques such as speedlines, flow ribbons, strobe silhouettes and opacity‐based techniques provide temporal context to the current timestep being visualized. We evaluated the effectiveness of these illustrative techniques by conducting a user study. We compared the ability of subjects to visually track features using snapshots, snapshots augmented by illustration techniques, animations, and animations augmented by illustration techniques. User accuracy, time required to perform a task, and user confidence were used as measures to evaluate the techniques. The results indicate that the use of illustration‐inspired techniques provides a significant improvement in user accuracy and the time required to complete the task. Subjects performed significantly better on each metric when using augmented animations as compared to augmented snapshots.

Visualization of off-screen data on tablets using context-providing bar graphs and scatter plots

Visualizing data on tablets is challenging due to the relatively small screen size and limited user interaction capabilities. Standard data visualization apps provide support for pinch-and-zoom and scrolling operations, but do not provide context for data that is off-screen. When exploring data on tablets, the user must be able to focus on a region of interest and quickly find interesting patterns in the data. We present visualization techniques that facilitate seamless interaction with the region of interest on a tablet using context-providing bar graphs and scatter plots. Through aggregation, fisheye-style, and overview+detail representations, we provide context to the users as they explore a region of interest. We evaluated the efficacy of our techniques with the standard, interactive bar graph and scatter plot applications on a tablet, and found that one of our bargraph visualizations - Fisheye-style Focus+Context visualization (BG2) resulted in the fewest errors, least frustration and took the least amount of time. Similarly, one of our scatter plot visualizations - User Driven Overview+Detail (SP3) - resulted in the fewest errors, least frustration and took the least amount of time. Overall, users preferred the context-providing techniques over traditional bar graphs and scatter plots, that include pinch-and-zoom and fling-based scrolling capabilities.

Unboxing cluster heatmaps

BackgroundCluster heatmaps are commonly used in biology and related fields to reveal hierarchical clusters in data matrices. This visualization technique has high data density and reveal clusters better than unordered heatmaps alone. However, cluster heatmaps have known issues making them both time consuming to use and prone to error. We hypothesize that visualization techniques without the rigid grid constraint of cluster heatmaps will perform better at clustering-related tasks.ResultsWe developed an approach to “unbox” the heatmap values and embed them directly in the hierarchical clustering results, allowing us to use standard hierarchical visualization techniques as alternatives to cluster heatmaps. We then tested our hypothesis by conducting a survey of 45 practitioners to determine how cluster heatmaps are used, prototyping alternatives to cluster heatmaps using pair analytics with a computational biologist, and evaluating those alternatives with hour-long interviews of 5 practitioners and an Amazon Mechanical Turk user study with approximately 200 participants. We found statistically significant performance differences for most clustering-related tasks, and in the number of perceived visual clusters. Visit git.io/vw0t3 for our results.ConclusionsThe optimal technique varied by task. However, gapmaps were preferred by the interviewed practitioners and outperformed or performed as well as cluster heatmaps for clustering-related tasks. Gapmaps are similar to cluster heatmaps, but relax the heatmap grid constraints by introducing gaps between rows and/or columns that are not closely clustered. Based on these results, we recommend users adopt gapmaps as an alternative to cluster heatmaps.

Visualizing disease incidence in the context of socioeconomic factors

Certain biological factors such as genetics, physical fitness, and lifestyle have been shown to influence an individuals risk of acquiring disease. But are there are other socioeconomic factors that influence disease incidence as well? In this paper, we introduce a visualization tool called Disease Trends that explores the associations and possible correlations between specific economic (personal income per capita), educational (percentage of adult population with a four year college degree), and environmental (air pollution level) factors with diabetes prevalence and cancer incidence rates across counties throughout the United States. It is structured as an interactive geographical visualization that displays disease incidence data as an interactive choropleth map and connects it with coordinated views of the socioeconomic variables for each county as the user scrolls over it. Additionally, the ability to compare and contrast counties as well as to interactively specify a region for comparison allows further examination of the data. This results in an informative overview of disease incidence trends that allows users to spot areas of interest and potentially pursue these areas further with more scientific research.