Patient-specific lean body mass can be estimated from limited-coverage computed tomography images

Nucl Med Commun. 2018 Jun;39(6):521-526. doi: 10.1097/MNM.0000000000000845.

Abstract

Objective: In PET/CT, quantitative evaluation of tumour metabolic activity is possible through standardized uptake values, usually normalized for body weight (BW) or lean body mass (LBM). Patient-specific LBM can be estimated from whole-body (WB) CT images. As most clinical indications only warrant PET/CT examinations covering head to midthigh, the aim of this study was to develop a simple and reliable method to estimate LBM from limited-coverage (LC) CT images and test its validity.

Patients and methods: Head-to-toe PET/CT examinations were retrospectively retrieved and semiautomatically segmented into tissue types based on thresholding of CT Hounsfield units. LC was obtained by omitting image slices. Image segmentation was validated on the WB CT examinations by comparing CT-estimated BW with actual BW, and LBM estimated from LC images were compared with LBM estimated from WB images. A direct method and an indirect method were developed and validated on an independent data set.

Results: Comparing LBM estimated from LC examinations with estimates from WB examinations (LBMWB) showed a significant but limited bias of 1.2 kg (direct method) and nonsignificant bias of 0.05 kg (indirect method).

Conclusion: This study demonstrates that LBM can be estimated from LC CT images with no significant difference from LBMWB.

MeSH terms

  • Adult
  • Aged
  • Aged, 80 and over
  • Body Weight*
  • Female
  • Humans
  • Image Processing, Computer-Assisted*
  • Male
  • Middle Aged
  • Positron Emission Tomography Computed Tomography*
  • Young Adult