Parametric response mapping of CT images provides early detection of local bone loss in a rat model of osteoporosis.

Loss of bone mass due to disease, such as osteoporosis and metastatic cancer to the bone, is a leading cause of orthopedic complications and hospitalization. Onset of bone loss resulting from disease increases the risk of incurring fractures and subsequent pain, increasing medical expenses while reducing quality of life. Although current standard CT-based protocols provide adequate prognostic information for assessing bone loss, many of the techniques for evaluating CT scans rely on measures based on whole-bone summary statistics. This reduces the sensitivity at identifying local regions of bone resorption, as well as formation. In this study, we evaluate the effectiveness of a voxel-based image post-processing technique, called the Parametric Response Map (PRM), for identifying local changes in bone mass in weight-bearing bones on CT scans using an established animal model of osteoporosis. Serial CT scans were evaluated weekly using PRM subsequent to ovariectomy or sham surgeries over the period of one month. For comparison, bone volume fraction and mineral density measurements were acquired and found to significantly differ between groups starting 3 weeks post-surgery. High resolution ex vivo measurements acquired four weeks post-surgery validated the extent of bone loss in the surgical groups. In contrast to standard methodologies for assessing bone loss, PRM results were capable of identifying local decreases in bone mineral by week 2, which were found to be significant between groups. This study concludes that PRM is able to detect changes in bone mineral with higher sensitivity and spatial differentiation than conventional techniques for evaluating CT scans, which may aid in clinical decision making for patients suffering from bone loss.