Automated three-dimensional measurement on anatomic parameters of proximal femur / 医用生物力学

ABSTRACT

ObjectiveTo automatically measure anatomic parameters of proximal femur by establishing three-dimensional (3D) coordinate system of the femur based on bony landmarks, so as to assist pre-operative planning and design of customized femoral stem. MethodsThe software named "Femeter" was independently developed for anatomic measurement of the femur, which allowed importing the femur models with STL format and manually locating the targeted anatomic landmarks. The 3D modeling of femoral medullary canal was rapidly created by semi-spherical iterative searching algorithm, and 16 key anatomic parameters of proximal femur, including femoral head radius, neck-shaft angle, anterversion were automatically calculated by least-squares fitting algorithm. ResultsBy experimenting on 30 femur STL models, the average execution time of single measurement was (0.95±0.16) seconds, and the intra-class correlation coefficient of 9 anatomic parameters was between 0.907 and 0.999, which showed high intra-rater and inter-rater reliability. ConclusionsThe automatic modeling and execution time of measuring algorithm by Femeter are fast, with satisfactory measurement repeatability and easy interaction, which is easily applicable to clinical practice. The 3D anatomic measurement of proximal femur can provide solid data for pre-operative planning of total hip arthroplasty, selection of commercialized femoral stem and design of customized femoral stem.