Alzheimer’s disease is a neurodegenerative disorder that is hard to clinically detect due its similar symptoms that are associated with a broad range of neurological disorders. However, due to the severity of the disorder it is important to detect differences in neuropathology early, and this can be achieved through the use of NeuroFormer, a model for analyzing data in neuroscience. EEG patterns between Alzheimer’s disease and even Frontotemporal dementia are subtle but distinct, so a sophisticated pattern recognition tool is necessary to detect differences in the frequency bands, temporal variations, and uniqueness within the neuropathologies. The NeuroFormer was tested using different learning rates, model structure, decision thresholds, and performance metrics to determine the best way of using NeuroFormer to detect Alzheimer’s. Ultimately, the NeuroFormer was found to have an accuracy of 95.76%, making it a superior model in analyzing and classifying Alzheimer’s disease. The authors conclude that the NeuroFormer is able to learn quickly, is effective for Alzheimer’s specific classification, has a balance between false alarms and missed cases in detecting Alzheimer’s and Frontotemporal dementia, as well as differences in using the model for screening and diagnosing patients. This is a proposed model for the use of NeuroFormer in Alzheimer’s detection, so future work that contains a larger population being tested or more data would ensure its ability to be applied to the field of medicine.