Fully automatic segmentation of femurs with medullary canal definition in high and in low resolution CT scans.

摘要: Abstract Femur segmentation can be an important tool in orthopedic surgical planning. However, order to overcome the need of experienced user with extensive knowledge on techniques, should fully automatic. In this paper a new automatic femur method for CT images is presented. This also able define automatically medullary canal and performs well even low resolution scans. Fully femoral was performed adapting template mesh volume medical images. achieve this, adaptation active shape model (ASM) technique based statistical (SSM) local appearance (LAM) novel initialization used, drive deformation fit in-image time effective approach. With proposed 98% convergence rate achieved. For high group average error less than 1 mm. image results are accurate 1.5 mm. The pipeline accurate, robust completely free. patient orientation, artifacts poorly defined edges. excelled significant slice thickness, i.e., above 5 mm. Medullary increases geometric information that used planning or finite element analysis.