Multi-Sensor Fusion of Biomedical Images

RIA的这项提议是开发使用多种方式组合心脏图像的技术：磁共振成像、正电子发射计算机断层扫描和CT。由于分辨率、大小和方向的不同，组合这类图像很困难。该技术是新颖的，它使用多层神经网络来完成图像处理任务。如果成功，该项目将改进分析生物医学图像的方法，从而潜在地改进医疗诊断。

This RIA proposal is to develop techniques to combine images of the heart using a variety of modalities: MRI, PET and CT. Combining such images is difficult because of differences in resolution, size and orientation. The proposed technique is novel; it uses multilayer neural networks for the image processing tasks. If successful, this project will result in improved methods of analyzing biomedical images, leading to potential improvements in medical diagnosis.