Abstract

Background and objectives: Bone cancer (sarcoma of bone) is a serious illness that happens when the healthy cells present in the bone change and become uncontrollable and form a mass called tumor. A bone tumor can be benign and malignant. Bone cancer broadly categorized as two types: Primary and Secondary where primary cancer starts from bone. Secondary bone cancer, initiate from the any parts of the body and then spread towards bone, it is called metastasis cancer. Bone cancer can be detected using a variety of imaging techniques, including scintigraphy, PET, MRI, CT scan, and X-rays. The purpose of this work is to disuses the complete process of bone cancer detection. This paper also discuss about the different approaches and techniques that can be used for preprocessing, segmentation and classification. Moreover, this work also compares different works and also discusses the major challenges and possibilities in this area. Material and Methods: The Bone MRI images are used for this work for Bone cancer detection. To perform the preprocessing on images ASF (Alternate sequential filtering), a Decision Median Filter (DEME-F) and an Unsymmetric Trimmed Median Filter (UNTME-F) is used. For the Augmentation different DI-Augmentation process is performed in terms of position augmentation as well as Intensity augmentation. For intensity augmentation brightness, contrast and saturation are boosted by using Enhanced grasshopper optimization algorithm. For Segmentation process, ASPP subsided and graded ASPP-steered segmentation technique (SubGrade ASPP) combined with Modified DeeplabV3+ (Mod2eep). Results: The segmented image is the output image.