When examining classifications related to scoliosis, "Idiopathic Scoliosis" emerges as the most prevalent type. Alongside spinal alterations, patients with scoliosis experience changes in stability and gait while standing. Although there are existing studies in the literature regarding the progression of scoliosis and its impact on plantar pressure among individuals diagnosed with adolescent idiopathic scoliosis, no studies have been found on predicting scoliosis risk in healthy adolescents. This study aims to develop a decision support system based on artificial neural networks (ANN) capable of predicting scoliosis risk in adolescents using foot pressure analysis values and machine learning models.
The study included 20 patients diagnosed with Adolescent Idiopathic Scoliosis and 43 healthy adolescent individuals with similar demographic characteristics (totaling 63 patients). Plantar pressure distributions of all participants were measured statically and dynamically.
Data collected for all patients included: age, gender, right hindfoot static plantar pressure percentage, left hindfoot static plantar pressure percentage, right forefoot static plantar pressure percentage, left forefoot static plantar pressure percentage, right foot dynamic plantar pressure percentage, and left foot dynamic plantar pressure percentage. A dataset was compiled with pressure percentages and the presence of scoliosis diagnosis information (comprising 8 input variables and 1 result variable for each patient).
The top performers in predicting adolescent idiopathic scoliosis risk were determined to be: Subspace KNN (100%), RUS Boosted Trees (100%), Weighted KNN (100%), Bagged Trees (100%), and Fine KNN (100%).
Primary Language | English |
---|---|
Subjects | Digital Health, Health Services and Systems (Other) |
Journal Section | Research Articles |
Authors | |
Publication Date | May 1, 2024 |
Submission Date | March 25, 2024 |
Acceptance Date | April 16, 2024 |
Published in Issue | Year 2024 Volume: 4 Issue: 1 |