Computer aided diagnosis (CAD) tools have the potential to support the radiologists in detection and can lead to improved diagnosis. This chapter presents deep learning based techniques for automated liver lesion analysis in computed tomography (CT) images. In the first part of this chapter we focus on the detection of liver metastases using both a global context with a fully convolutional network (FCN), and a local patch level analysis with a superpixel sparse based classification. In the second part of this chapter we present a novel system for the generation of virtual PET images using CT scans. We combine an FCN with a conditional generative adversarial network (GAN) to generate simulated PET data from given input CT data and show its benefit by reducing the false-positive rate.
|Title of host publication||Handbook of Medical Image Computing and Computer Assisted Intervention|
|Number of pages||26|
|State||Published - 1 Jan 2019|
- Conditional GAN
- Liver Lesions