Isabel Harb Manssour

Visualizing inner structures in multimodal volume data

With the evolution of medical image acquisition techniques, the capacity and fidelity of image-based diagnosis were extended. The current trend is to acquire information using multiple sources to help medical diagnosis, but the integration of the multivariate data into a single 3D representation is non-trivial. Techniques for the visualization of multimodal volume data have been developed with the goal of finding suitable strategies to integrate characteristics of multiple data sets into a single visual representation. Likewise, several techniques are dedicated to the exploration of different ways of incorporating seeing-through capabilities into volume rendering techniques. This paper presents a new approach to visualize inner structures in multimodal volume data, which is based in the utilization of cutting tools.

A framework to visualize and interact with multimodal medical images

The simultaneous use of images obtained from different sources is common in medical diagnosis. However, even though the quality of these images has been improving, the integration of multimodality data into a unique 3D representation is still non-trivial. To overcome this problem, multimodal visualization techniques provide better insight by finding suitable strategies to integrate different characteristics of multiple data sets into a single visual representation. This paper describes a framework for interactive multimodal visualization of 3D medical images, focusing on the multimodal visualization model and requirements for developing such systems. A short overview of multimodal visualization systems and techniques is also presented.

Exploring Twitter Interactions through Visualization Techniques: Users Impressions and New Possibilities

Social networking websites have been taking a huge space on the Internet. With millions of members spread across the world, these services connect their users through “friendship” relations, and within a short period of time the information is disseminated in the network. Considering Twitter, we investigated how standard techniques of information visualization can improve the analysis of its users’ behavior, i.e. the group of people who most closely relate, and the frequency and type of messages they send. This investigation was done through user’s studies with a set of visualization techniques implemented in an interactive system prototype.

A novel physics-based interaction model for free document layout

Marketing flyers, greeting cards, brochures and similar materials are expensive to produce, since these documents need to be personalized and typically require a graphic design professional to create. Either authoring tools are too complex to use or a predefined set of fixed templates is available, which can be restrictive and difficult to produce the desired results. Thus, simpler design tools are a compelling need for small businesses and consumers. This paper describes an interactive authoring method for creating free-form documents based on a force-directed approach, traditionally applied for graph layout problems. This is used for automatically distributing and manipulating images, text and decorative elements on a page, according to forces modeled after physical laws. Such approach can be used for enabling easy authoring of personalized brochures, photo albums, calendars, greeting cards and other free-form documents. A prototype has been developed for evaluation purposes, and is briefly described in this paper. Evaluation results are presented as well, showing that users enjoy the experience of designing a page by interacting with it, and that end results can be satisfactory.

Augmented reality environment for life support training

The area of Medical Qualification in Life Support training is being constantly improved. However, many problems still have to be faced in the training sessions. During these sessions the students or physicians can repetitively practice patient care procedures in simulated scenarios using anatomical manikins, especially designed for this type of training. Current manikins have several resources incorporated to allow and facilitate qualified training, such as pulse, arrhythmia and auscultation simulator. However, some deficiencies have been detected in the existing LS training structure. For example: automatic feedback to the students in consequence of their actions on the manikin, images like facial expressions and body injuries, and their combination with sounds that represent the clinical state of the patient. The main goal of the ARLIST project is to qualify the traditional training environment currently used for LS training, introducing image and sound resources into the training manikins. Thought these features we can simulate some aspects such as facial expressions, skin color changes and scratches and skin injuries through image projection over the manikin body, and also play sounds like cries of pain or groans of an injured man.

High performance approach for inner structures visualisation in medical data

The capacity and fidelity of image-based diagnosis were extended due to the evolution of medical image acquisition techniques. Such data is usually visualised through volume rendering, which denotes a set of techniques used to present three-dimensional images with the main goal of showing the interior of a volume and enabling the identification of its inner regions and structures. Several tools described in the literature are dedicated to explore different ways of incorporating seeing-through capabilities into volume rendering techniques. However, these tools are based on visualisation algorithms and are usually computationally intensive, especially when working with large datasets. An alternative to optimise rendering time is to use high performance programming to implement such tools, thus providing a faster response to user interaction. This paper presents a new approach to visualise inner structures in medical volume data using a parallel ray casting algorithm to allow user interaction with the volume.

A multimodal visualization framework for medical data

Although the use of information acquired from multiple sources to help in medical diagnosis is increasing, the integration of multivariate data into a unique 3D representation is non-trivial. Multimodal visualization techniques have been proposed with the goal of finding suitable strategies to integrate important characteristics of multiple data sets providing better insight. Considering the requirements of multimodal visualization systems, this work describes a framework for interactive multimodal visualization of volume data.

Crosswalk Localization from Low Resolution Satellite Images to Assist Visually Impaired People

In this article, we propose a model for crosswalk detection and localization by using satellite images captured from Google Maps, for the purpose of assisting visually impaired people. The detection is performed by an SVM classifier, which is combined with Google Road Map to speed up computation time and to eliminate some possible false alarms. We assume that a visually impaired person holds a smartphone with an embedded GPS, which is used to initialize the extraction of images from Google Maps, as well as to assist its user by providing audio feedback of the nearest detected crosswalk. This issue brings forward significant interest and it is also very challenging, mainly due to illumination changes, occlusion, image noise and resolution, besides the quality of crosswalks that sometimes are badly painted in many developing countries. Experimental results indicate that the proposed model works well in low resolution images, effectively detecting and localizing crosswalks in simulated scenarios.

Towards a Domain-Specific Language for geospatial data visualization maps with Big Data sets

Data visualization is an alternative for representing information and helping people gain faster insights. However, the programming/creating of a visualization for large data sets is still a challenging task for users with low-level of software development knowledge. Our goal is to increase the productivity of experts who are familiar with the application domain. Therefore, we proposed an external Domain-Specific Language (DSL) that allows massive input of raw data and provides a small dictionary with suitable data visualization keywords. Also, we implemented it to support efficient data filtering operations and generate HTML or Javascript output code files (using Google Maps API). To measure the potential of our DSL, we evaluated four types of geospatial data visualization maps with four different technologies. The experiment results demonstrated a productivity gain when compared to the traditional way of implementing (e.g., Google Maps API, OpenLayers, and Leaflet), and efficient algorithm implementation.

Two new aesthetic measures for item alignment

This paper introduces two methods for measuring the alignment of items on a page with respect to its left/right margins. The methods are based on the path followed by the eyes as they follow the items from top to bottom of the page. Examples are presented and both methods are analyzed with respect to the axioms presented in [2], that describe how good alignment measure is supposed to behave.