Figure 1 | Scientific Reports

Figure 1

From: Deep learning-based patient re-identification is able to exploit the biometric nature of medical chest X-ray data

Figure 1

General problem scenario: Comparing a given chest radiograph to publicly available dataset images by means of DL techniques would either result in discrete labels indicating whether or not the dataset images belong to the same patient as the given radiograph (verification scenario) or yield a ranked list of the most similar radiographs related to the given scan (retrieval scenario). Images belonging to the same patient are highlighted with the same color. The given radiograph is marked with an asterisk. The shown cases would enable a potential attacker to link sensitive patient-related information contained in the dataset to the image of interest.

Back to article page