Early Detection of Guillain-Barre Syndrome through Machine Learning Using Nerve Conduction Study

Guillain-Barre Syndrome (GBS) is an extremely rare neurological condition that needs to be diagnosed and treated as soon as possible. The Nerve Conduction Study (NCS) is a frequently performed diagnostic test that offers insightful information on peripheral nerve activity. This research aims to investigate the feasibility to diagnose GBS by analyzing NCS reports and applying machine learning techniques. We have collected data from 163 subjects at the National Institute of Neurosciences and Hospital, which contains both afflicted and unaffected individuals. Later, the diagnosis of GBS patients was made using machine learning algorithms. The total diagnosis was made using four different machine learning algorithms: support vector machines (SVM), K nearest neighbour (KNN), logistic regression and random forest. The accuracies of the diagnosis processes were calculated and the highest accuracy of 93% was found using the random forest and logistic regression model. Evaluation measures such as accuracy, precision, recall, and Fl-score are also computed for verifying the effectiveness of the models. The findings of this study illustrate the potential of machine learning approaches in diagnosing GBS utilizing NCS reports. The study can be helpful for medical doctors in diagnosing the GBS patients more effectively.