Facial Skin Disorder Prediction Based on Non-Visual Information Using ANN Model

Abstract

Facial skin disorders are common health problems affecting a person’s quality of life. While diagnosing facial skin disorders usually requires a direct visual examination by a dermatologist, non-visual information can sometimes help the diagnosis process. Currently, facial skin disorders such as acne and hyperpigmentation are the dominant disorders that are most often the problem of every patient who visits the clinic for treatment, according to experts. This study aims to apply an ANN (Artificial Neural Networks) model that can predict facial skin disorders using non-visual data. Age, gender, skin type, family history, symptoms, and risk factors are non-visual data used to train the ANN model. A method used to accurately predict various facial skin disorders using non-visual information integrated into an ANN model. Feature selection, model building, training, validation, and model performance evaluation enable the identification of facial skin disorders based on non-visual information. The results show that ANN technology and non-visual information can offer overall diagnosis efficiency to predict facial skin disorders. Furthermore, the accuracy obtained based on non-visual data is 100%.

Access to Document

Other files and links

• Link to publication in Scopus

Fingerprint