08 Mar Dr Molton University Research Projects in Cosmetic Medicine
In a recent research paper for the IEEE Access Journal, Dr Molton of Epiclinic explores how machine learning can be used as a facial rejuvenation prediction tool.
Dr Molton shows how real and synthetic data can be used to predict the outcome of facial rejuvenation before a cosmetic procedure is even carried out. This can be achieved by first mapping out the underlying structures of the face using 3D imaging technology, which will then allow doctors to estimate how much dermal filler will need to be applied and what the outcome will be.
Rejuv3DNet is the proposed technology that Epiclinic will use to make these predictions. To add to this, Dr Molton has also developed the first 3D face cosmetic data set, which comprises real pre and post-treatment 3D images as well as synthetically-generated 3D images of the face.
To date, experimental results show that the technology is able to achieve a 62.5% and 66.67% prediction accuracy using real-world data, while synthetic data sets produced a 75.2% and 89.5%, and 77.2% and 90.1% prediction accuracy.
These reported accuracies will be used as proof of concept, while additional data will be used to make further improvements.
This innovative approach to cosmetic medicine has great potential and will definitely warrant further investigation.