Sensitivity analysis of paediatric knee kinematics to the graft surgical parameters during anterior cruciate ligament reconstruction: a sequentially linked neuromusculoskeletal-finite element analysis
Ayda Karimi Dastgerdi,Amir Esrafilian,Christopher P. Carty,Azadeh Nasseri,Martina Barzan,Rami K. Korhonen,Ivan Astori,Wayne Hall,David John Saxby
DOI: https://doi.org/10.1016/j.cmpb.2024.108132
IF: 6.1
2024-03-12
Computer Methods and Programs in Biomedicine
Abstract:Background and objective Incidence of paediatric anterior cruciate ligament (ACL) ruptures has increased substantially over recent decades. Following ACL rupture, ACL reconstruction (ACLR) surgery is performed to restore knee function. This surgery involves replacing the failed ACL with a graft, however, surgeons must select from range of surgical parameters (e.g., type, size, insertion, and pre-tension) with no robust evidence guiding these decisions. Methods This study used sequentially-linked neuromusculoskeletal (NMSK) finite element (FE) models of three paediatric knees to estimate the: (i) sensitivity of post-operative knee kinematics to four surgical parameters (type, size, insertion, and pre-tension) through multi-input multi-output sensitivity analysis; (ii) influence of different motion and loading conditions throughout stance phase of walking gait on sensitivity indices; and (iii) influence of subject-specific anatomy (i.e., knee size) on sensitivivty indices. Additionally, a validated FE model of the intact knee for each subject served as a reference against which we compared ACLR knee kinematics. Results Sensitivity analyses revealed significant influences of surgical parameters (i.e., graft type, size, insertion, and pre-tension) on ACLR knee kinematics, albeit without discernible trend favouring any one parameter. Graft size and pre-tension were primary drivers of variation in knee translations and rotations, however, their effects fluctuated across stance indicating motion and loading conditions affect system sensitivity to surgical parameters. Importantly, the sensitivity of knee kinematics to surgical parameter varied across subjects, indicating knee geometry (i.e., gross size) influenced system sensitivity. Notably, alterations in graft parameters yielded substantial effects on kinematics (normalized root-mean-square-error > 10%) compared to intact knee models, indicating surgical parameters vary post-operative knee kinematics. Conclusions Overall, this initial study highlights the importance of surgical parameter selection on post-operative kinematics in the paediatric ACLR knee, and provide evidence of the need for personalized surgical planning to ultimately enhance patient outcomes.
engineering, biomedical,computer science, interdisciplinary applications,medical informatics, theory & methods