Smart System to Detect Brain Age Using Machine Learning

Abstract

Detecting brain ageing is of paramount importance in the medical field due to its significant contribution to the rising number of deaths each year. Brain ageing stands out as a prevalent health concern, characterized by a high mortality rate and widespread occurrence. Extensive research endeavors are underway to address this issue, with Magnetic Resonance Imaging (MRI) emerging as a pivotal tool for identifying and tracking the progression of brain ageing. MRI scans offer detailed insights into the ageing process, facilitating superior outcomes compared to alternative methodologies. In our paper, we propose an innovative approach for detecting brain ageing using MRI scanned images. The methodology encompasses several crucial steps, beginning with image preprocessing, where the application of a median filter enhances image quality. Subsequently, segmentation techniques employing mathematical morphological operations isolate regions indicative of brain ageing. Geometric features such as area, perimeter, and eccentricity are then computed for the identified ageing regions. The culmination of our approach involves the utilization of an Iterative Convolutional Neural Network (CNN) classifier. This classifier distinguishes between ageingous (malignant) and normal (benign) brain regions based on the extracted features. To further enhance the accuracy of our classification, we employ both Artificial Neural Network (ANN) as a baseline method and introduce the Optimistic Convolutional Neural Network (OCNN), a novel algorithm proposed in our research. Through rigorous experimentation and evaluation, we compare the performance of ANN and OCNN, analyzing their respective accuracies. Our findings unequivocally demonstrate that the OCNN outperforms the traditional ANN, offering superior accuracy and efficacy in detecting brain ageing from MRI scans. This underscores the potential of advanced neural network architectures in revolutionizing medical image analysis and diagnosis. In conclusion, paper presents a robust methodology for detecting brain ageing using MRI scanned images, leveraging state-of-the-art image processing techniques and innovative neural network algorithms. By enhancing the accuracy and efficiency of brain ageing detection, our research contributes significantly to the ongoing efforts aimed at mitigating the adverse impacts of this pervasive health issue.