GANs can be used to generate deceivingly real looking radiographs and to merge databases of radiological images without disclosing patient sensitive information. This contributes to solving the important problem of building large radiological image databases for the training of computer vision algorithms. GANs that have been trained in-house may serve as a mean to distribute the information contained within the database without actually providing a real snapshot of patient sensitive data: only the weight distribution of the GAN needs to be transferred and a representative artificial dataset of millions of radiographs may be generated in reasonable computational time at a peripheral site. GANs can be used to visualize what the generator neural network sees and to reveal correlations between diseases. This might be used in several ways: as a check that the GAN has been trained correctly, as a tool to uncover relationships between diseases or to visualize hallmark changes of pathologies and potentially also as a decision support system for diagnosis.
Input variables : real database of radiographs
Output Variables : Artificial Radiographs
Visit Model : github.com
Additional links : biorxiv.org
| Model Category | : | Public |
| Date Published | : | November, 2019 |
| Healthcare Domain | : | Medical Technology |
| Code | : | github.com |
| Data Privacy |
| Synthetic Data Generation |