The generation of three-dimensional (3D) medical images can have great application potential since it takes into account the 3D anatomical structure. There are two problems, however, that prevent effective training of a 3D medical generative model: (1) 3D medical images are very expensive to acquire and annotate, resulting in an insufficient number of training images, (2) a large number of parameters are involved in 3D convolution. To address both problems, we propose a novel GAN model called 3D Split&Shuffle-GAN. In order to address the 3D data scarcity issue, we first pre-train a two-dimensional (2D) GAN model using abundant image slices and inflate the 2D convolution weights to improve initialization of the 3D GAN. Novel 3D network architectures are proposed for both the generator and discriminator of the GAN model to significantly reduce the number of parameters while maintaining the quality of image generation. A number of weight inflation strategies and parameter-efficient 3D architectures are investigated. Experiments on both heart (Stanford AIMI Coronary Calcium) and brain (Alzheimer's Disease Neuroimaging Initiative) datasets demonstrate that the proposed approach leads to improved 3D images generation quality with significantly fewer parameters.
翻译:三维(3D)医学图象的生成具有巨大的应用潜力,因为它考虑到了 3D 解剖结构。然而,有两个问题妨碍了3D医学基因模型的有效培训:(1) 3D医学图象的获取和注释费用非常昂贵,导致培训图象数量不足,(2) 3D演化涉及大量参数。为了解决这两个问题,我们提议了一个名为 3D Split & Shuffle-GAN的新型GAN模型。为了解决3D数据稀缺问题,我们首先用丰富的图像切片预入二维(2D)GAN模型,并用2D演化的重力来更新2D的GAN模型,以改进3DGAN模型的初始化,从而导致培训图象数量不足; 提议为GAN模型的生成者和导师建议建立大量3D网络结构,以大幅降低参数数量,同时保持图像生成的质量。 调查了一些重量通胀策略和参数效率3D结构。为了解决3D数据稀缺的问题,我们先用丰富的图像切片预入两维(2D) GAN模型,用丰富的图像切片切片切片和大脑(Ahemamaims legages dismaismaismas) leging the dismaismaisgages progages palgismismations