Through aging and injury, the intervertebral disc of the lumbar spine can undergo degeneration, leading to collapse of the vertebrae and low back pain, a symptom that affects half the adult population in any given year. In an effort to reduce low back pain, total disc replacement treatment removes the degenerated disc, restores natural height and lordosis of the segment, and preserves motion at the joint. Patient anatomy, implant selection, and implant placement play significant roles in a patient’s outcomes after total disc replacement surgery. Thus, the objective of the work presented in this thesis was to develop a suite of statistical and computational tools describe population-based anatomy and to support component selection and placement in TDR surgical procedures with the goal of improving implant design and patient outcomes.

The statistical modeling approach quantified shape and alignment variation of the lumbar spine by characterizing variability of shape and size of individual vertebra, relative alignment of relevant segments, and overall anatomy of the lumbar spine. Statistical shape models of single vertebrae revealed that the primary mode of variation correlated to vertebral body size variation (average R² = 0.82 across vertebrae), which can inform sizing lines for total disc replacements. Strong correlations of disc height to the second (R² = 0.82) and third (R² = 0.88) principal components of the shape-alignment models of the L4-L5 and L5-S1 segments are useful in assisting clinicians diagnose pathologies, screening patients for treatment options, and pre-operatively planning for surgical treatment. Statistical models of the entire spine reveal how vertebral shape changes influence the spine as a whole.

The subject-specific templating approach of total disc replacement surgeries accurately predicted ROM in a cohort of twenty two patients implanted with the ProDisc-L device and suggested changes to total disc replacement size selection and alignment to improve ROM. Predicted ROM was 11.8% different to actual ROM. Improvements in ROM could have been achieved in over 85% of the cases had the proposed templating process been employed, which showed that pre-operative templating can be an important tool to achieve maximum ROM and optimal clinical outcomes.

Computational pilot evaluations of subjects implanted with the Activ-L device provided insight into the mechanical behavior of a total disc replacement featuring a center inlay that can translate within the inferior end plate. Results indicated that greater translation of the inlay related to greater overall ROM. Subjects implanted with the Activ-L achieved greater ideal range of motion than subjects with a ProDisc-L, a device featuring an inlay that is fixed within the inferior end plate. Further investigations into this work can reveal design considerations that significantly influence ROM and patient outcomes.