Auto-Segmentation of Lumbar Vertebrae using MultiPlanar UNet

Vertebra segmentation is a fundamental step for spinal image analysis and interference. The lumbar backbone is critical for transferring load and permissive drive, but identifying spinal irregularities and fractures is possibly disputing. Automated medical images can assist doctors in diagnosing these environments more expertly than manual or semi-manual procedures. To advance the discovery and assessment of compression fractures utilizing MRI figures, a deep learning approach is suggested. MRI figures pose one-of-a-kind challenges due to coinciding forms and hazy boundaries, with added determinants. We propose a deep learning-based approach utilizing the MP U-Net model for segmenting individual vertebrae in MRI figures of the lumbar backbone. The ending step engages in fine-tuning the segmentation by changing the hyperparameters. This approach offers advantages over classical procedures and can potentially enhance the analysis and treatment of spinal settings.