Personalised trainer recommendation based on physical activity and genetic profile

Foot-care specialists recommend shoes by analysing the patient’s gait cycle and looking for any structural or functional problems. Such methods are time consuming, inaccurate and unable to identify any risk factors that may lead to development of foot-related diseases in the future. This work presents a footwear recommendation algorithm based on genetic predispositions i.e. the genetic profile associated to selected Single Nucleotide Polymorphisms (SNPs), and the individual activity level, in addition to age, body mass index (BMI) and pronation. The algorithm, built on an Artificial Neural Network (ANN), returns a personalised recommendation for four different commercially available shoe categories (Minimalist, Stability, Motion Control, Cushioned). The activity profiles are generated based on features extracted from actual users’ step count data collected via the wearable device DnaBand™, which are then combined with users’ physical information and genetic profile. The Gaussian Mixture Model (GMM) has been found to best identify the relevant activity profiles’ clusters. 5 case studies have been selected and used to validate the ANN output.