Artificial intelligence-enabled simulation of gluteal augmentation: A helpful tool in preoperative outcome simulation?

Abstract

While the buttock region is considered an esthetic hallmark, the Brazilian butt lift (BBL) remains controversially discussed in the plastic surgery community. This is due to its contentious safety profile. Thus, informed consent and patient education play a key role in preoperative planning. To this end, we aimed to program an easy-to-use, widely accessible, and low-budget algorithm that produces reliable outcome simulations. The conditional generative adversarial network (GAN) was trained using pre- and postoperative images from 1628 BBL patients. To validate outcome simulation, 25 GAN-generated images were assessed deploying 67 Amazon Mechanical Turk Workers (Mturks). Mturks could not differentiate between GAN-generated and real patient images in approximately 49.4% of all trials. This study presents a free-to-use, widely accessible, and reliable algorithm to visualize potential surgical outcomes that could potentially be applied in other fields of plastic surgery.