Automatic Bone Cancer Detection: An Extensive Survey

Abstract

Bone cancer is a range of complexities and diversities of malignancies which stem from bone itself (primary bone cancer) or from other parts of the body (secondary bone cancer). The early diagnosis and proper identification of bone cancer is essential to proper treatment and patient care. This survey paper covers the wide range of bone cancer detection techniques covering both primary and secondary bone cancer. It describes the different types of primary bone cancer (osteosarcoma, Ewing sarcoma, chondrosarcoma and chordoma), their clinical features, prevalence and the problems with their diagnosis. In addition, secondary bone cancer is discussed, with particular emphasis on the most common primary cancers that secondarily involve the bones – breast cancer, lung cancer, prostate cancer, and kidney cancer. Conventional imaging techniques, emerging imaging techniques, biomarkers, radiomics, machine learning techniques and genetic markers are explored in detail. The current challenges, progress and future outlook in bone cancer detection are also discussed. It is aimed at providing a valuable resource for researchers, clinicians and healthcare professionals in the diagnosis and management of bone cancer with the aim of improving early detection, making a more accurate diagnosis and improving the individual approach to treatment.