Measurement of changes in trabecular bone at fracture sites using X-ray CT and automated image registration and...

Currently, there are no in vivo techniques for quantifying healing within fractures involving mainly trabecular bone. Methods for quantitatively assessing healing would aid the investigation of new treatment regimes, and might also be useful for predicting whether a patient may be undergoing delayed union. Using CT and automated image registration, we have developed an image processing technique for measuring changes in CT image intensity at fracture sites. The technique focuses on quantifying the formation of new mineralised tissue within fracture gaps, while ignoring loss of bone mineral due to disuse osteoporosis. Seven patients with fractures of the distal radius were examined for 12 weeks following fracture. To assess reproducibility of measurements of change in CT intensity at the fracture line, measurements were performed on two separate occasions, by each of two independent readers. Reproducibility was compared to rates of change over time, to determine detectable differences in individual progression. Scans were scored qualitatively for features of healing and scores compared to the quantitative measurements. The mean (SD) change in CT intensity was +128 (65) Hounsfield Units (HU) over the 12 week follow-up. Inter- and intra-observer reproducibilities were both similar (+/- 17 HU), 1/7 of the change seen during the study. In this small patient cohort, a significant increase in CT intensity was seen 2 weeks post-fracture. Large early increases in CT intensity were associated with early visual appearance of sclerosis and blurring of the fracture line. In this preliminary, prospective study, we have developed a reproducible quantitative technique for measuring changes in CT intensity of trabecular bone at the fracture line in the distal radius. Further work is required to determine whether it can be used to identify, or monitor patients who are undergoing delayed fracture repair. The technique appears sensitive for measuring changes immediately post-fracture, and could have a role in examining potential effects of new therapies in patient cohorts.