Image segmentation and registration algorithm to collect homologous landmarks for age-related thoracic morphometric...

Adults 65 years of age and older currently constitute more than 12% of the total population and the elderly population is projected to reach nearly 20% by 2030. Previous studies have shown that skeletal and physiological resilience decline with age, resulting in a decreased ability for the body to withstand traumatic insults. In the current study, an algorithm was developed to quantify age and gender-specific variations in the thoracic skeletal morphology. Normal chest CT scans of males and females ages 0-100 were collected from a radiological database. Image segmentation and subsequent image registration was used to collect landmark data from the ribs. Rigid, affine, and non-rigid, non-linear transformations are used to morph segmented ribs from different subjects to a “rib atlas”. The atlas consists of a normal chest CT scan from an average male with over 100 landmarks placed per rib. The transformation matrices are used to map landmarks from the atlas coordinate system to the coordinate system of each CT scan, effectively allowing for collection of homologous (or comparable) rib landmarks across all subjects. Geometric morphometrics will be used in future work to analyze the landmark data to formulate age and gender-specific shape and size variation functions. These functions will be used to create a scalable finite element model of the thorax that will be used to predict thoracic injury response for different ages and genders.