Automatic Detection of lumbar Disc herniation using Computer Vision and Artificial Intelligence

During the last few years, deep learning-based methods have set new standards for many computer vision and pattern recognition research. Computer vision, as a core technology of Artificial Intelligence, has been one of the most compelling advancements of the 21st century. While there have been notable difficulties in the technology ecosystem of computer vision, medical image segmentation has emerged as a particularly notable application. The goal of medical image segmentation is to provide a precise and accurate representation of the object of interest within the image, typically for diagnosis, treatment planning, and qualitative analysis. However, due to the complex nature of most medical images, producing quality ground truth data can be challenging due to factors such as complex backgrounds, noise, and human error. In this project, we propose a method of automatically generating ground truth data from MRI images through instance segmentation, using a proposed accepted confidence and consistent metric to measure the resulting ground truth data's quality. We leverage an Artificial intelligence-based system to annotate the region of interest in the MRI image and utilize Mask RCNN, a deep neural network architecture, with a mean average precision of 98%, which can localize and specify the location of the disc in the image. And afterwards classifying the Region of interest with an accuracy of 70%. Our approach aims to help radiologists automatically detect the region of interest in MRI Images, leading to easier diagnoses with certainty of quality ground truth data, The contributions of this study include a novel approach to generating ground truth data for medical image segmentation and the use of advanced technologies such as deep neural networks to automate the process. The potential benefits of this approach include more reliable and accurate analysis of medical images, leading to improved patient outcomes.