Class-Aware Adversarial Pulmonary Nodule Synthesis

An apparatus, comprising:

• means for receiving 1) an initial medical image patch having a masked portion and an unmasked portion, and 2) a class label associated with a nodule to be synthesized;
• means for generating a synthesized medical image patch using a trained generative adversarial network, the synthesized medical image patch comprising the unmasked portion of the initial medical image patch and a synthesized nodule replacing the masked portion of the initial medical image patch, the synthesized nodule being synthesized according to the class label;
• and
• means for outputting the synthesized medical image patch.

A non-transitory computer readable medium storing computer program instructions, the computer program instructions when executed by a processor cause the processor to perform operations comprising:

• receiving 1) an initial medical image patch having a masked portion and an unmasked portion, and 2) a class label associated with a nodule to be synthesized;
• generating a synthesized medical image patch using a trained generative adversarial network, the synthesized medical image patch comprising the unmasked portion of the initial medical image patch and a synthesized nodule replacing the masked portion of the initial medical image patch, the synthesized nodule being synthesized according to the class label;
• and
• outputting the synthesized medical image patch.