• Deepika . B
• Vishnupriya . J

There is a significant increase in the number ofcases related to different spine pathologies in the recent years. For an accurate computer aided diagnostic system, an unambiguous and reliable segmentation of the human vertebral column is crucial. It aids doctors in detecting different spine pathologies with improved visualizations. Magnetic Resonance Imaging (MRI) is the best method available to visualize spine. MR images of spine are clearer and more detailed than images obtained from other imaging methods. Correct segmentation of vertebrae from MR images is a challenging task because of the high variations in shape topology, low contrast of images and different fields of view. We propose an efficient computer aided diagnostic system for fully automatic segmentation of vertebrae. Our approachuses Mask R-CNN, a deep neural network algorithm that was initially developed to detect objects and perform instance segmentation of objects in natural images. In our paper, we demonstrate the application of Mask R-CNN in segmentation of vertebrae in Magnetic Resonance images of spine. The aim of this system is only to perform segmentation and a medical professional is required to arrive at an accurate diagnosis from thesegmentation results.

Medical Image Segmentation, Mask R-CNN, Vertebrae Segmentation, MRI Spine

"Medical Image Segmentation Application", International Journal of Emerging Technologies and Innovative Research (www.jetir.org), ISSN:2349-5162, Vol.7, Issue 5, page no.85-89, May-2020, Available :http://www.jetir.org/papers/JETIRDV06020.pdf

"Medical Image Segmentation Application", International Journal of Emerging Technologies and Innovative Research (www.jetir.org | UGC and issn Approved), ISSN:2349-5162, Vol.7, Issue 5, page no. pp85-89, May-2020, Available at : http://www.jetir.org/papers/JETIRDV06020.pdf