Literature DB >> 33428062

Synthetic CT image generation of shape-controlled lung cancer using semi-conditional InfoGAN and its applicability for type classification.

Ryo Toda1, Atsushi Teramoto2, Masakazu Tsujimoto3, Hiroshi Toyama4, Kazuyoshi Imaizumi4, Kuniaki Saito1, Hiroshi Fujita5.   

Abstract

PURPOSE: In recent years, convolutional neural network (CNN), an artificial intelligence technology with superior image recognition, has become increasingly popular and frequently used for classification tasks in medical imaging. However, the amount of labelled data available for classifying medical images is often significantly less than that of natural images, and the handling of rare diseases is often challenging. To overcome these problems, data augmentation has been performed using generative adversarial networks (GANs). However, conventional GAN cannot effectively handle the various shapes of tumours because it randomly generates images. In this study, we introduced semi-conditional InfoGAN, which enables some labels to be added to InfoGAN, for the generation of shape-controlled tumour images. InfoGAN is a derived model of GAN, and it can represent object features in images without any label.
METHODS: Chest computed tomography images of 66 patients diagnosed with three histological types of lung cancer (adenocarcinoma, squamous cell carcinoma, and small cell lung cancer) were used for analysis. To investigate the applicability of the generated images, we classified the histological types of lung cancer using a CNN that was pre-trained with the generated images.
RESULTS: As a result of the training, InfoGAN was possible to generate images that controlled the diameters of each lesion and the presence or absence of the chest wall. The classification accuracy of the pre-trained CNN was 57.7%, which was higher than that of the CNN trained only with real images (34.2%), thereby suggesting the potential of image generation.
CONCLUSION: The applicability of semi-conditional InfoGAN for feature learning and representation in medical images was demonstrated in this study. InfoGAN can perform constant feature learning and generate images with a variety of shapes using a small dataset.

Entities:  

Keywords:  CNN; CT imaging; Classification; GAN; Image synthesis; Lung cancer

Mesh:

Year:  2021        PMID: 33428062     DOI: 10.1007/s11548-021-02308-1

Source DB:  PubMed          Journal:  Int J Comput Assist Radiol Surg        ISSN: 1861-6410            Impact factor:   2.924


  2 in total

Review 1.  Early detection of lung cancer: clinical perspectives of recent advances in biology and radiology.

Authors:  F R Hirsch; W A Franklin; A F Gazdar; P A Bunn
Journal:  Clin Cancer Res       Date:  2001-01       Impact factor: 12.531

2.  Automated Pulmonary Nodule Classification in Computed Tomography Images Using a Deep Convolutional Neural Network Trained by Generative Adversarial Networks.

Authors:  Yuya Onishi; Atsushi Teramoto; Masakazu Tsujimoto; Tetsuya Tsukamoto; Kuniaki Saito; Hiroshi Toyama; Kazuyoshi Imaizumi; Hiroshi Fujita
Journal:  Biomed Res Int       Date:  2019-01-02       Impact factor: 3.411

  2 in total
  2 in total

1.  MS-ResNet: disease-specific survival prediction using longitudinal CT images and clinical data.

Authors:  Jiahao Han; Ning Xiao; Wanting Yang; Shichao Luo; Jun Zhao; Yan Qiang; Suman Chaudhary; Juanjuan Zhao
Journal:  Int J Comput Assist Radiol Surg       Date:  2022-04-20       Impact factor: 3.421

2.  Lung cancer CT image generation from a free-form sketch using style-based pix2pix for data augmentation.

Authors:  Ryo Toda; Atsushi Teramoto; Masashi Kondo; Kazuyoshi Imaizumi; Kuniaki Saito; Hiroshi Fujita
Journal:  Sci Rep       Date:  2022-07-27       Impact factor: 4.996

  2 in total

北京卡尤迪生物科技股份有限公司 © 2022-2023.