Abstract: Alzheimer’s disease (AD), which is a neuro associated disease, has become a common for past few years. In this competitive world, individual has to perform lot of multi tasking to prove their efficiency, in this process the neurons in the brain gets affected after a while i.e., “Alzheimer’s Disease”. Existing models to identify the disease at early stage has taken the individuals speech as input then they are converted into textual transcripts. These transcripts are analyzed using neural network approached by integrating them with NLP techniques. These techniques failed in designing the model which can process the long conversation text at faster rate and few models are unable to recognize the replacement of the unknown words during the translation process. The proposed system addresses these issues by converting the speech obtained into image format and then the output “Mel-spectrum” is passed as input to pre-trained VGG-16. This process has greatly reduced the pre-processing step and improved the efficiency of the system with less kernel size architecture. The speech to image translation mechanism has improved accuracy when compared to speech to text translators.

Ragavamsi Davuluri and Ragupathy Rengaswamy, “A Pre-trained Neural Network to Predict Alzheimer’s Disease at an Early Stage” International Journal of Advanced Computer Science and Applications(IJACSA), 13(5), 2022. http://dx.doi.org/10.14569/IJACSA.2022.0130524

@article{Davuluri2022,
title = {A Pre-trained Neural Network to Predict Alzheimer’s Disease at an Early Stage},
journal = {International Journal of Advanced Computer Science and Applications},
doi = {10.14569/IJACSA.2022.0130524},
url = {http://dx.doi.org/10.14569/IJACSA.2022.0130524},
year = {2022},
publisher = {The Science and Information Organization},
volume = {13},
number = {5},
author = {Ragavamsi Davuluri and Ragupathy Rengaswamy}
}