• POOJA N
• AISHWARYA S
• SHASHIKALA K S
• REVATHI S A

Segmentation of knee bones from MRI has advanced as a tool for the analysis of knee joint pathologies. This paper intended at creating a fully automated bone segmentation model for knee bone and cartilages such as femur and tibia by using annotated MR images. Knee MR Images are pre-processed, augmented, annotated and segmented using Neural Networks. The first step in the proposed model is preprocessing and augmentation where cropping, resizing and resampling occurs. SimpleITK toolkit is a library which provides certain readymade functions for preprocessing and segmentation. In the next step we create annotated images using 3D Slicer. Annotation helps in detecting the object (knee bone and cartilages) from the image. Convolutional neural network method called U-Net is used as the fundamental method to perform segmentation of the knee bones and cartilage. Segmented bones are used for the Age Assessment using the ossification degree of growth plates. Age Prediction is a convoluted process where determining the chronological oldness of a person who lacks the authorized certification is done. While no technique provides a proper method to identify the age, the proposed system helps in analyzing the age using the growth plates present in the bone. This paper provides a step towards a solution by fully automating the extraction of bone in knee MRIs using convolutional neural networks to reduce the data complexity. Furthermore, the discussions concerning the MRI classification results used to segment the bone will be discussed.

Magnetic resonance image, Annotation, U-Net, Age assessment, Augmentation

"AUTOMATIC SEGMENTATION OF BONE AND CARTILAGES USING ANNOTATED KNEE IMAGES", International Journal of Emerging Technologies and Innovative Research (www.jetir.org), ISSN:2349-5162, Vol.7, Issue 5, page no.197-202, May-2020, Available :http://www.jetir.org/papers/JETIR2005167.pdf

"AUTOMATIC SEGMENTATION OF BONE AND CARTILAGES USING ANNOTATED KNEE IMAGES", International Journal of Emerging Technologies and Innovative Research (www.jetir.org | UGC and issn Approved), ISSN:2349-5162, Vol.7, Issue 5, page no. pp197-202, May-2020, Available at : http://www.jetir.org/papers/JETIR2005167.pdf