Determination of tissue volumes - A comparison between CT and MR imaging

Purpose. To evaluate MR imaging for tissue volume measurements. Material and Methods: Imaging was done with a spoiled gradient echo techniq ue using the human lower leg as a simple model and CT as the reference tech nique. Areas of adipose tissue, muscle plus skin and bone tissue were evalu ated in 5 volunteers using signal intervals (MR) and attenuation intervals (CT) defined by histograms. Volumes were determined from 7 images (MR-7, CT -7), from 70 contiguous images (MR-70), and also from a commercially-availa ble 3D program based on a seeding technique applied according to the operat ors' judgment (MR-op). Results. The sum of bone tissue, muscle plus skin and adipose tissue were s imilar for CT and MR. Compared to CT-7, MR-7 and MR-70 overestimated adipos e tissue (0-8%) and muscle plus skin (11-15%), while bone tissue was undere stimated (61-78%). Determinations based on the contiguous images or the 7 i mages were similar (standard error 2-5%). MR-op resulted in pronounced unde restimation of muscle plus skin and bone tissue and a minor underestimation of adipose tissue. Conclusion: Valid determinations of adipose tissue volume are possible with MR but further technical development is required for other tissue volumes. The technique applied according to the operators' judgement, as used for 3 D surface rendition, is unreliable for tissue volume determinations.