A Modified UNet based Framework towards Early Detection of Autism using EEG Waves

Autism is a spectrum disorder with a multitude of brain abnormalities if not detected at the earlier stages, it leads to major complications. It’s a kind of neurological disorder. Many methods are available in the literature to detect autism such as imaging modalities like MRI, PET, CT, and EEG. EEG is a measure to record brain signals and autism is detected in it by analyzing the EEG wave spectrogram. The proposed methodology incorporates Deep Learning based method to detect autism through an EEG image spectrogram. It has given about 98% of accuracy as compared to other classical approaches.