Asynchronous QCT for Volumetric Bone Density Assessment

Asynchronous QCT was performed in baseline CT, a technique that provides results comparable to conventional QCT [26 (link)]. Attenuation values in HU were manually sampled with tools of the institutional picture archiving and communication system software (Sectra IDS7, Sectra AB) and transformed into volumetric BMD with conversion equations calculated by asynchronous calibration. An experienced radiologist placed a circular region of interest in trabecular bone of lumbar vertebrae L1 to L4, as previously described [27 ], using on-the-fly calculated midsagittal stacks of 15-mm thickness. Sampled HU was averaged over assessed vertebrae, omitting fractured vertebra or those with apparent alterations of the trabecular bone due to degeneration or hemangioma.
HU-to-BMD conversion equations were calculated by linear regression, in three scanners (Philips Brilliance 64, iCT 256, and Siemens Somatom Definition AS+) based on measurements of density-reference phantoms (QRM) in dedicated scans with the same tube voltage and scanner settings as in clinical routine acquisitions, and in two already decommissioned scanners (Siemens Somatom Definition AS and Sensation Cardiac 64) based on retrospective measurements of a density-reference phantom (Osteo Phantom, Siemens Healthineers), which had been included in the scanner couch during clinical CT scans for a certain period of time in the past (Fig. 2). Retrospective measurements of the Siemens Osteo phantom and a second calibration phantom (Mindways Software) were performed in CT exams, which were randomly selected from the institutional database in 2-month intervals over the entire time period when phantoms were present. Thereby, long-term scanner stability was evaluated in three scanners (Philips iCT 256, Siemens Somaton Definition AS, and Sensation Cardiac 64). Conversion equations and long-term stability measures are shown in Table 4. A BMD correction offset for contrast-enhanced CT scans with arterial (− 8.6 mg/cm³) and portal venous contrast phase (− 15.8 mg/cm³) was added based on previous investigations [28 (link)]. Osteoporosis was defined as BMD < 80 mg/cm³ and osteopenia as 80 ≤ BMD ≤ 120 mg/cm³ [29 ].Fig. 2

Routine CT scan of a 63-year-old female patient for follow-up purpose after metastatic gastric cancer and liver transplant with administration of oral and intravenous contrast medium in portal venous phase. For two MDCT scanners (Siemens Somatom Definition AS [in this example] and Sensation Cardiac 64), retrospective measurements of an in-plane calibration phantom present underneath patients during routine scans were used for asynchronous calibration and evaluation of long-term scanner stability