Accuracy Prediction of the Type of Primary Headache and Prevent Misdiagnosis of Headaches Using Support Vector Machine (SVM)

Abstract

Primary headaches can be broadly classified as Cluster Headaches, Migraine, Tension-Type Headache (TTH)[56] , and Primary Stabbing-type headache[25] . A headache is classified by the symptoms associated with it. The frequency,  duration,  severity of pain,  information on the time of day when the headache occurred,  any kind  of sensory sensitivity such as light; use of prescribed or non-prescribed medicines; hours of sleep obtained at night; emotional state; weather and/or daily activity; dietary consumption during the last twenty four hours; and any known underlying health conditions and accompanying autonomic symptoms at that time and family history are decisive factors to diagnosing the type of headache and providing a suitable solution[6,41] . However, there have been several cases of misdiagnosis of headache(s) due to little or no information of the cause or trigger of the headache(s), the patient history and the conditions under which it occurs. The aim is to prevent the misdiagnosis of the type of headache by developing SVM-based diagnostic tool by considering parameters such as duration, autonomic symptoms, the region of pain, pain severity, light sensitivity, frequency, and temporal pattern.