An AI model developed by researchers at Erasmus MC can predict the risk of skin cancer based on a facial photo better than the current methods used by doctors.
The team trained the AI model with more than 2,800 facial images, all taken under similar conditions, where the researchers were able to identify which individuals went on to develop skin cancer.
200 characteristics of skin cancer
From these photos, the algorithm extracted 200 features to predict the risk of developing skin cancer, both on the face and elsewhere on the body. The AI model can even identify risk factors for skin cancer that are invisible to the human eye.
More accurate
The predictions of the AI model appear to be more accurate than predictions based on traditional methods such as skin research, questionnaires and genetic research. These findings were published in eClinicalMedicine.
Further research
Although the AI model is not yet ready for clinical use, the focus is on further research. The researchers want to test the algorithm on larger datasets with photos of people of different ages and ethnic backgrounds. In addition, consideration is being given to sharing the model publicly so that fellow scientists can build on it further.
Tags: model surpasses existing methods predicting skin cancer risk
