Non-Fungible Tokens Using Machine Learning Simulation: An Intertwining of Two Digital Technologies for the Aesthetic Surgeon

Goals/Purpose: Modern plastic surgeons rely on media, usually in the form of before-and-after photographs, to exhibit their surgical results. With the explosion of social media as well as low-cost photo manipulation software, there is an increasing risk of two digital dangers: the fraudulent manipulation of images to falsely elevate a surgeon’s results, and the theft of a surgeon’s legitimate intellectual property to be shared on another’s outlets. Surgeons attempt to mitigate the latter by using image watermarks to ensure proper attribution. However, defending the provenance of these images is paramount, and now enabled with non-fungible blockchain technology. Additionally, valid simulation attempts by the surgeon are complex and expensive, lacking in artificial intelligence and possibly not based on true clinical outcomes. Combining accurate simulations with proof of the legitimacy of the results enables the creation of a digital fingerprint that safeguards the surgeon and provides confidence to the patient.

Methods/Technique: A sample breast augmentation case performed by a surgeon in an ASAPS accredited fellowship was selected. Two modern technologies were applied. First, standard preoperative photographs were analyzed by an artificial intelligence trained through hundreds of iterations of breast augmentation simulations to provide generative modeling. The projected postoperative model was positioned between the true pre and postoperative photographs. A non-fungible token (NFT) was minted utilizing the image combination.

Results/Complications: The postoperative photograph closely resembles the simulated result, attesting to the validity of the machine learning platform. The image trio was successfully minted using Ethereum blockchain technology, creating an indelible digital record of validated transactions and attributable to the surgeon.

Conclusion: Here we present the marriage of two novel technologies: machine learning simulation of surgical results and minting of non-fungible tokens utilizing pre and postoperative images. The combination can usher in a modern standard for both accurate prediction and the authenticity of the aesthetic outcome, providing automatic attribution to the surgeon and his/her intellectual property. This digital fusion creates a transparent record of the proposed and actual results for any viewer to observe.