2020/10/07 by David Chettrit, Tomer Meir, Chettrit, David +10
Engineering · Medicine · #Bone and Joint Diseases #Computer Vision and Pattern Recognition (cs.CV) #FOS: Computer and information sciences #FOS: Electrical engineering #Image and Video Processing (eess.IV) #Medical Imaging and Analysis #Spinal Fractures and Fixation Techniques #electronic engineering #information engineering
paper · pdf · doi:10.48550/arxiv.2010.03739
openalex publication_date 2020/10/07 · openalex created_date 2022/07/25 · openalex updated_date 2026/07/28
An osteoporosis-related fracture occurs every three seconds worldwide,\naffecting one in three women and one in five men aged over 50. The early\ndetection of at-risk patients facilitates effective and well-evidenced\npreventative interventions, reducing the incidence of major osteoporotic\nfractures. In this study, we present an automatic system for identification of\nvertebral compression fractures on Computed Tomography images, which are often\nan undiagnosed precursor to major osteoporosis-related fractures. The system\nintegrates a compact 3D representation of the spine, utilizing a Convolutional\nNeural Network (CNN) for spinal cord detection and a novel end-to-end sequence\nto sequence 3D architecture. We evaluate several model variants that exploit\ndifferent representation and classification approaches and present a framework\ncombining an ensemble of models that achieves state of the art results,\nvalidated on a large data set, with a patient-level fracture identification of\n0.955 Area Under the Curve (AUC). The system proposed has the potential to\nsupport osteoporosis clinical management, improve treatment pathways, and to\nchange the course of one of the most burdensome diseases of our generation.\n