MEDICAL IMAGE PRESENTATION

我们的计划继续调查医疗表现的手段， 图像来自广泛的成像模式。 研究，包括 六个项目，两个核心，涉及技术和临床应用 3D和2D显示用于诊断和放射治疗计划。 对物体定义和人类视觉的研究开发了模型， 支持显示技术发展的方法。 在3D显示领域，1）我们将开发3D勘探技术， 医学图像数据和用于可视化的复杂组合， 解剖和规划对象，以及2）我们将研究3D显示 在放射治疗和病理学中的应用。 在我们的2D显示工作中，1）我们 将评估乳房X光检查对比度改善的诊断用途， 放射治疗门户网站，和CT胸部图像，和B）我们将开发和 在现场和现场评估临床可用的诊断工作站 实验室，用于多模态和乳房X线摄影显示。 我们的 人类视觉项目和我们的物体定义项目将侧重于形状， 以及复杂语境下基于群体的感知与分析 在医学图像中发现，并将产生形状的视觉模型。 的 对象定义项目将包括交互式3D应用程序 放射治疗射野图像的显示和自动在线分析。 将在所有项目领域进行有控制的评价。 一个统计核心不仅将提供设计和数据分析， 观察员实验，但也发展改进的统计或 在具有相关观测的小样本的观测者研究中进行分析， 在这两种情况下，对可感知性进行评级，并给出连续的响应。 研究结果将允许较小的样本提供保证类型 错误率和统计功效。 一个设施核心将支持软件和硬件设施， 管理和人体工程学咨询。

Our program continues to investigate means of presentation of medical images from a wide range of imaging modalities. The research, involving six projects and two cores, concerns techniques and clinical applications of 3D and 2D display for diagnosis and the planning of radiotherapy. Investigations into object definition and human vision develop models and methods supporting display technique development. In the 3D display area, 1) we will develop techniques for 3D exploration of medical image data and for visualization of complicated combinations of anatomy and planning objects, and 2) we will investigate 3D display applications in radiotherapy and pathology. In our 2D display work, 1) we will evaluate diagnostic uses of contrast improvements of mammographic, radiotherapy portal, and CT chest images, and b) we will develop and evaluate clinically usable diagnostic workstations in the field and in the laboratory, for both multimodality and mammographic display. Both our human vision project and our object definition project will focus on shape- and grouping-based perception and analysis in the complicated contexts found in medical images and will produce models of vision of shape. The object definition project will include application in both interactive 3D display and automatic, on-line analysis of radiotherapy portal images. Controlled evaluations will be conducted in all of the project areas. A statistics core will not only provide designs and data analyses for observer experiments but also development of improved statistics or analyses in observer studies with small samples of correlated observations, both where perceivability is rated and where continuous responses are give. The research results will allow smaller samples to provide guaranteed type I error rates and statistical power. A facilities core will support software and hardware facilities, administration, and ergonomic consultation.