Research on Feature Selection and Predicting ALS Disease Progression

Abstract

Amyotrophic lateral sclerosis (ALS) is a rapidly progressive, invariably fatal neurological disease that attacks the nerve cells responsible for controlling voluntary muscles. The disease belongs to a group of disorders known as motor neuron diseases, which are characterized by the gradual degeneration and death of motor neurons. Although ALS is incurable and fatal, with median survival of 3–5 years, treatment can extend the length and meaningful quality of life for patients. Here, to be useful clinically, we tried several feature selection methods to choose predictive features identified using ALS clinical trials dataset. The feature selection method of random frog coupled with partial least square is an exact way that can be helpful for predictive feature selection. We further apply the proposed regression method partial least square regression to predict 3–12 month ALS progression slope, as measured using the ALS functional rating scale (ALSFRS). The experiment results show that the proposed selector and predictor has shown itself to be robust to extreme outliers. It is of great benefit to accelerate ALS research and development, identify new disease predictors and potentially significantly reduce the costs of future ALS clinical trials.