Dual Channel Convolutional Neural Network for Dementia Disease Prediction

Dementia disease is a complex neurological condition that is affecting millions of population around the world. Dementia is known as a umbrella term because it does not represent a specific disease, it describes a range of cognitive impairments. Which are interfered significantly with the individual's ability to function independently in their daily life. As a person's age increases, the risk of dementia development also increases. This disease primarily affects the memory, thinking, and communication skills. The traditional methods failed to diagnose the disease accurately in early stage. This paper proposes a Dual Channel Convolutional Neural Network (DCCNN) to predict and classify the dementia disease. This algorithm is employed with the two different convolution layers that are internally consists of three convolutional layers and three fully connected layers and these layers are trained with the dataset first, and that are capable of processing two different type of data or features that are in that particular dataset simultaneously by utilizing two separate pathways or channels that are within the same model or architecture. Further, this method combines the features and provides the accurate prediction of the dementia. The proposed DCCNN model obtained 93% for R-fMRI dataset compared to the existing methods.