Correlation via synthesis: end-to-end nodule image generation and radiogenomic map learning based on generative adversarial network
This work addresses the need for noninvasive identification of molecular properties in diseases like lung cancer, but it is incremental as it builds on existing GAN methods for medical image synthesis.
The authors tackled the problem of learning radiogenomic maps linking image features to gene expression profiles by developing an end-to-end generative adversarial network (GAN) that synthesizes images conditioned on both background images and gene data, tested on a non-small cell lung cancer dataset, resulting in realistic synthetic images and a promising approach for holistic gene-image relationship discovery.
Radiogenomic map linking image features and gene expression profiles is useful for noninvasively identifying molecular properties of a particular type of disease. Conventionally, such map is produced in three separate steps: 1) gene-clustering to "metagenes", 2) image feature extraction, and 3) statistical correlation between metagenes and image features. Each step is independently performed and relies on arbitrary measurements. In this work, we investigate the potential of an end-to-end method fusing gene data with image features to generate synthetic image and learn radiogenomic map simultaneously. To achieve this goal, we develop a generative adversarial network (GAN) conditioned on both background images and gene expression profiles, synthesizing the corresponding image. Image and gene features are fused at different scales to ensure the realism and quality of the synthesized image. We tested our method on non-small cell lung cancer (NSCLC) dataset. Results demonstrate that the proposed method produces realistic synthetic images, and provides a promising way to find gene-image relationship in a holistic end-to-end manner.