DOI

10.5703/1288284317871

Description

Neurodegenerative disorders can cause signal disturbances on the neuronal spiking patterns within the brain. Parkinson’s disease, as an example, produces burst firings and synchronized oscillatory signal patterns. Accurately distinguishing between different brain spiking pattern activities is essential in finding effective neuromodulation or neurostimulation treatments for neurodegenerative disorders like Parkinson disease. Due to the lack of publicly available neuronal spiking pattern datasets containing different neuronal signal patterns and classes, we propose the generation of a novel three-class dataset. We then convert the generated signals into 2D images and separate the generated signal images into test and train datasets. We aim to design, implement, train, and evaluate custom neural network architectures to classify the generated signals. After developing five distinct neural network models, we conducted a comparative analysis of their performance. This work represents a giant leap towards real-time neuronal signal classification and the detection of the abnormalities in brain neuronal activity

Share

COinS
 

Neuronal Signal Activity Image Classification

Neurodegenerative disorders can cause signal disturbances on the neuronal spiking patterns within the brain. Parkinson’s disease, as an example, produces burst firings and synchronized oscillatory signal patterns. Accurately distinguishing between different brain spiking pattern activities is essential in finding effective neuromodulation or neurostimulation treatments for neurodegenerative disorders like Parkinson disease. Due to the lack of publicly available neuronal spiking pattern datasets containing different neuronal signal patterns and classes, we propose the generation of a novel three-class dataset. We then convert the generated signals into 2D images and separate the generated signal images into test and train datasets. We aim to design, implement, train, and evaluate custom neural network architectures to classify the generated signals. After developing five distinct neural network models, we conducted a comparative analysis of their performance. This work represents a giant leap towards real-time neuronal signal classification and the detection of the abnormalities in brain neuronal activity