Abstract
Privacy concerns around sharing personally identifiable information are a major barrier to data sharing in medical research. In many cases, researchers have no interest in a particular individual’s information but rather aim to derive insights at the level of cohorts. Here, we utilise generative adversarial networks (GANs) to create medical imaging datasets consisting entirely of synthetic patient data. The synthetic images ideally have, in aggregate, similar statistical properties to those of a source dataset but do not contain sensitive personal information. We assess the quality of synthetic data generated by two GAN models for chest radiographs with 14 radiology findings and brain computed tomography (CT) scans with six types of intracranial haemorrhages. We measure the synthetic image quality by the performance difference of predictive models trained on either the synthetic or the real dataset. We find that synthetic data performance disproportionately benefits from a reduced number of classes. Our benchmark also indicates that at low numbers of samples per class, label overfitting effects start to dominate GAN training. We conducted a reader study in which trained radiologists discriminate between synthetic and real images. In accordance with our benchmark results, the classification accuracy of radiologists improves with an increasing resolution. Our study offers valuable guidelines and outlines practical conditions under which insights derived from synthetic images are similar to those that would have been derived from real data. Our results indicate that synthetic data sharing may be an attractive alternative to sharing real patient-level data in the right setting.
Highlights
Sharing sensitive data under strict privacy regulations remains a crucial challenge in advancing medical research[1]
We benchmarked the generation of synthetic medical image data to closely mimic the distribution-level statistical properties of a real source dataset
We evaluated two state-of-the-art generative adversarial networks (GANs) models, prog-GAN and cpDGAN, on two real-world medical image corpora consisting of chest radiographs and brain computed tomography (CT), respectively
Summary
Sharing sensitive data under strict privacy regulations remains a crucial challenge in advancing medical research[1]. By accessing large amounts of collected data, there have been impressive research results in a range of medical fields such as genetics[2], radiomics[3,4], neuroscience[5], diagnosis[6,7,8], patient outcome prediction[9,10] or drug discovery[11,12]. It is especially difficult to share and distribute medical data due to privacy concerns and the potential abuse of personal information[15]. To overcome these privacy concerns there has been an impressive number of large-scale research collaborations to pool and curate de-identified medical data for opensource research purposes[16,17,18]. Privacy laws inhibit medical data sharing[20], and potentially available de-identification methods lack guarantees as de-identified data can, in some cases, be linked back to individuals[21,22]
Sign-in/Register to view highlights & summary Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callDisclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
