Bones are designed to withstand tremendous amounts of force, including the forces incurred by walking and running. The femur, the large bone in the upper leg also known as the thigh bone, articulates or connects with the pelvic bone forming the hip joint, and with the tibia and kneecap forming the knee joint. Despite their strength, bones can break if subjected to excessive forces or if they are not well nourished. A fractured bone can cause pain, limping, tenderness, swelling or deformity. It can be detected by taking an X-ray. A X-ray can also show whether a broken bone is out of alignment and if it has healed properly.
X-rays can also show the condition of other tissues and organs. The femur bone can be affected by diseases of the blood, muscles, joints and the bones themselves. A femur X-ray can also detect cysts, tumors and other abnormalities. In addition, a femur X-ray can help diagnose conditions that affect the articular surface of the bone, such as arthritis and osteochondrosis.
The tibia is the lower leg bone, which extends from the knee to the ankle. It is the strongest bone in the human body. The tibia can be affected by injuries such as ligament damage, broken bones or sprains. A tibia injury may also be caused by a fall, car accident or other trauma. An X-ray can detect the presence of a fracture, whether it is open or closed.
A long bone has a tube-like shaft called the diaphysis. It contains the proximal and distal ends of the bone and is filled with yellow bone marrow in an adult. The ends of the diaphysis meet at a wider section near the middle of the bone, called the epiphysis. The epiphysis is covered by a plate of spongy bone called the metaphyseal plate that contains vascular channels and allows for growth during childhood and adolescence.
The present work has developed and evaluated an automated image stitching method to produce long limb X-ray panorama images using a minimal number of sector image acquisitions. The procedure utilises the medial and lateral femur and tibia edges captured in the fixed and moving image sectors, minimising the distance between the homologous bone edge pairs in both sets of images. It was found that this approach yielded comparable results to the expert manual reconstruction of a full-length panorama. This is especially evident when examining the comparison of the minimum translations obtained by the automated algorithm (yellow ‘X’) and the values selected by the expert manually (green). The evaluation was performed on a collection of 95 patient images acquired as part of a multi-site clinical trial. This collection served as the benchmark dataset for testing the effectiveness of this approach. See Figure 2 for a demonstration of the image stitching process.