Computer Vision News - November 2022

32 AI in Orthopedics and it is often lower in cost. However, X-ray lacks the spatial information available in CT, therefore, it is challenging to reconstruct a 3D model with comparable accuracy. To address the challenge of 3D reconstruction from standard X-ray images, both the unique AI technology, including dedicated deep neural networks, Total knee arthroplasty (TKA) is a common orthopedic procedure, conducted when the knee joint malfunctions and cannot be repaired. Classic open surgery resulted in large incisions, and success depended on the surgeon’s experience. New technological improvements brought assistant robots into the operation room (OR) and democratized TKA. Most robotic assisted surgeries (RAS) include a planning stage, where a 3D model of the knee is acquired and used for prosthesis assessment approach planning. This model is produced using dedicated algorithms from a knee CT scan, as it provides a high- resolution 3D scan, resulting in an accurate model. RSIP Vision has recently begun developing a solution for reconstructing an accurate 3D model of the knee from two X-ray images, anterior-posterior (AP) and lateral . Creating such a solution obviates the need for a CT scan, benefitting the procedure from two aspects - clinically and operationally. Standard X-ray imaging results in significantly less harmful radiation exposure to the patient compared to a CT scan, ultimately reducing procedural risk. Additionally, X-ray imaging has better accessibility than CT, it is more widely reimbursed in the U.S. healthcare system, 2D-TO-3D KNEE BONES RECONSTRUCTION - SUCCESSFUL PRELIMINARY RESULTS FROM CLINICAL STUDY