Intelligent System for Analyzing MRI Images to Find the Main Elements of the Human Spine

Abstract

This study aims to build an intelligent system based on MRI image analysis to find the main elements of the human spine, such as vertebrae and intervertebral discs. The core of this intelligent system is to design a neural network model that can accurately identify and find vertebrae and intervertebral discs. Therefore, this paper proposes an improved YOLOv8-seg image segmentation model, which introduces two lightweight convolution modules, GhostConv and C3Ghost, to reduce the computational complexity and parameter complexity of the model while maintaining high performance; at the same time, the CBAM attention mechanism is introduced to focus on specific features to improve model performance. The results show that the intelligent system built based on the improved YOLOv8-seg image segmentation model can automatically find and segment the vertebrae and intervertebral disc areas in lumbar MRI images, with high accuracy and recall, significantly improving segmentation efficiency and reducing human errors. In the future, we will further optimize the model structure and increase the diversity and scale of the data set to further improve the performance and stability of the intelligent system.