Authors:
Divya Saxena, Sukaant Chaudhary

Addresses:
1Knight Foundation School of Computing and Information Sciences, Florida International University, Miami, Florida, USA. 2eCloud Labs Inc, Iseline, New Jersey, United States of America. dsaxe001@fiu.edu1, sukaant@gmail.com2 

Abstract:

FMRI classification with neural networks entails training a machine learning model to categorize brain pictures collected from MRI scans. This method uses neural networks to learn complicated patterns and features from MRI data, allowing for accurate and automated classification of brain disorders or structures. Various neural network architectures have been explored and evaluated to determine the most effective model for distinguishing between fMRI images of individuals who are healthy and those who are patients. A promising new method for identifying brain illnesses and examining brain structure is fMRI classification with neural networks. Although this method is still being refined, it has the potential to completely alter how we now identify and treat brain illnesses. The job determines which neural network design should be used for fMRI categorization. For instance, a CNN may be a viable option if the objective is to categorize fMRI data from a particular time point. However, an RNN or LSTM network may be preferable if the objective is to categorize fMRI data from a series of time points. MRI images are a valuable source of information for diagnosing brain illnesses. However, manually classifying MRI images is a time-consuming and challenging task. This is where neural networks come in. Neural networks can be used to automatically classify MRI images with high accuracy. Various combinations of neural networks have been explored to classify MRI images to identify brain illness. In addition to individual neural networks, there have also been some studies that have explored the use of hybrid neural networks. Neural network categorization for MRI is a rapidly developing field. As neural networks get stronger, they can classify MRI pictures more accurately and with more accuracy.

Keywords: Magnetic Resonance Imaging (MRI); Convolutional Neural Network (CNN); Graph Convolution Networks (GCN); Recurrent Neural Network (RNN); Long Short-Term Memory (LSTM); Neuro-Imaging; functional MRI (fMRI); Machine Learning.

Received on: 17/10/2022, Revised on: 29/12/2022, Accepted on: 27/01/2023, Published on: 19/02/2023

FMDB Transactions on Sustainable Computer Letters, 2023 Vol. 1 No. 1, Pages: 33-48

• Views : 185
• Downloads : 10

Download PDF