Ultrasound and motion capture analysis for pre-operative planning in lower limb joint replacement surgeries

Pre-operative planning in total knee and hip arthroplasty is important for surgical outcome and patient satisfaction. Current clinical gold standards for pre-operative planning include imaging methods which are invasive to the patient and limited to one position of analysis. Lower limb and pelvic alignment are assessed in planning for total knee and hip arthroplasty respectively and have shown to vary in their measurements between standing and supine. B-mode ultrasound has shown to be a promising method for gaining superficial structures like muscles and bones. B-mode ultrasound can be performed rapidly and is relatively cheap and measurements can be conducted with the patient in various positions. The aim of this thesis is to establish non-invasive protocols for pre-operative planning in knee and hip surgeries.
Several approaches were developed to non-invasively measure lower limb and pelvic alignment. These consisted of using integrated motion capture and ultrasound system (OrthoPilot, Aesculap). A smart system (Aesculap) which consisted of a smart phone, smart tablet and ultrasound device was used to measure pelvic tilt from the anterior pelvic plane. A motion capture system on its own was used to measure the pelvic tilt in alternative manners. And finally, a synchronised ultrasound and motion capture setup was used for three-dimensional reconstructions of bone geometries. Supine and standing measurements were conducted which showed the flexibility of the measurements unlike common alternatives (X-Ray, MRI, CT).
Several operators performed precise measurements of key lower limb parameters. For example, varus-valgus was shown to be measured within 1 degree across operators. Femur and tibia segment lengths were also consistent (<5mm maximum variation between operators). Femur and tibia torsion measurements were less reliable (up to 10-15 degrees of variation between operators). Pelvic tilt measurements were also found to be unreliable regardless of the measurement technique. Initial promise and feasibility of three-dimensional reconstructions of all lower limb joint axis for implementation into musculoskeletal models was also shown. Joint contact forces differences between the implementation of MRI and ultrasound parameters into the models were less than 1 body weight.
Overall, ultrasound has shown to be useful in the assessment of lower limb parameters and bone geometries. This work has built upon previous findings to continue its development in the field of pre-operative planning and musculoskeletal modelling. Further work will include a large validation of subject-specific musculoskeletal modelling from ultrasound reconstructions. Improvements to the lower limb assessment with OrthoPilot will also be investigated.