Medical image privacy preservation via image synthesis and filtration

A computer-implemented method, comprising:

• accessing, by at least one of one or more devices operatively coupled to a processor, a medical image on which a first artificial neural network has been executed;
• training, by at least one of the one or more devices, a second artificial neural network on the medical image to perform image synthesis;
• generating, by at least one of the one or more devices and via execution of the second artificial neural network post-training, a set of synthetic variants of the medical image;
• and
• instructing, by at least one of the one or more devices, a computerized device to fine-tune the first artificial neural network using at least some of the set of synthetic variants of the medical image rather than using the medical image.

A computer program product for facilitating medical image privacy preservation via image synthesis and filtration, the computer program product comprising a non-transitory computer-readable memory having program instructions embodied therewith, the program instructions executable by a processor to cause the processor to:

• access a privacy-restricted medical image which a clinical machine learning model has analyzed;
• generate, via execution of a generative model that is trained on the privacy-restricted medical image to perform image synthesis, a set of unrestricted synthetic variants of the privacy-restricted medical image;
• and
• re-train the clinical machine learning model using at least some of the set of unrestricted synthetic variants instead of the privacy-restricted medical image.