Global Cancer Statistics 2020 report noticed over three hundred eight thousand new incidence cases and two hundred and fifty thousand new deaths were happened across the globe in 2020 alone due to Brain and nervous system disorders. Hence, early-stage identification and diagnosing of Brain and nervous system disorder is very important. Migraine is a neurological illness that causes severe headaches and has a substantial impact on patients’ lives. Specialists may find it difficult and time-consuming to diagnose Migraine Disease. As a result, systems that can assist professionals in the early detection of migraine disease is essential. Although migraine is one of the most frequent neurological illnesses, there have been very few studies on migraine diagnosis, particularly those using electroencephalogram (EEG) and deep learning (DL). The proposed method has three main phases. Data conversion from signal to spectrogram images, design of customized CNN and training and testing using various networks. The performance of the proposed-net for detection of migraine has significantly higher than the pretrained networks of Alexnet, Resnet and VGG16.