A Comparative Analysis of Stimuli Response among People with Migraine Classification: A Machine Learning Approach

Recurrent, throbbing headaches are the hallmark of the neurological condition known as migraines. Other symptoms that are frequently present with migraines include nausea, vomiting, and light and sound sensitivity. This work developed a machine learning model that can almost perfectly identify migraine by analysing electroencephalography (EEG) signals. The data set used in this study consists of EEG signals recorded from 21 healthy and 18 migraine patients and obtained in three different stimuli: resting state, visual stimulus and auditory stimulus. From a total of 144 channels, 14 channels were considered as the data source. The filtered raw EEG signals were converted into raw amplitude representation within a data frame. We used a total of 3 cases for this comparative analysis by taking a single stimuli at a time, a two-stimuli hybrid approach and all three stimuli combined approach. After the basic preprocessing steps were applied, the data were fed into some Machine learning algorithms. Among the algorithms, Logistic Regression with a two-stimuli hybrid approach showed 99.74% accuracy. This approach, marked by its high accuracy, may pave the way for a reliable point-of-care solution in the diagnosis and management of migraine, offering a potential breakthrough in enhancing patient care and treatment strategies within the realm of neurological disorders. This proposed lightweight machine learning model can be effective for less calculation time and less cost.