5q-Spinal muscular atrophy (SMA) is a rare disease that can now be treated, presenting new challenges for pediatricians as treatment decisions evolve with disease-modifying therapies.
A data-driven machine learning approach was used to analyze a dataset of 84 SMA patients, focusing on predicting scoliosis by selecting relevant features through expert knowledge and statistical methods.
The study identified key predictors like motor function scores, age, weight, and various medical interventions, achieving a prediction accuracy of 82% using a Random Forest Classifier.