Validation of a densitometric calibration protocol in microCT

Method. A group of 30 healthy trabecular bone specimens were selected to validate the densitometric calibration protocol. Specimens were ashed in a muffle furnace at 650 degrees for 24 hours. The ash density (ρASH) was calculated as the ash weight divided by the specimen volume and was found to range between 200 and 500 mg/cm. The ρASH is considered the golden standard for densitometric analysis, therefore was used in order to validate the micro computed tomography (microCT) calibration protocol. The densitometric calibration protocol in microCT was based on the use of a distilled water pet and two densitometric phantoms (250 and 750 mg/cm hydroxyapatite (HA) bone-equivalent density). Acquisition settings for both water and densitometric phantoms were 50kV, 200μA and 5.9s exposure time for each frame. The final image was obtained by averaging two frames, as previously published [1, 2]. An 1-mm Aluminium (Al) filter was used to reduce beam hardening effect. The rotation step was set at 0.90° to reduce the time of the acquisition. The magnification was set at 16x in order to obtain a pixel size of 19.5μm and a field of view of 20x20mm. The densitometric phantoms were dipped in saline to reproduce the same experimental set-up of trabecular bone specimens. The two phantoms were acquired together in the same holder. The tomographic acquisitions of distilled water and phantoms were reconstructed in 8-bit format with the beam hardening correction implemented in the reconstruction software. The beam hardening correction was set at 30%. The reconstructed slices of the water and the densitometric phantoms were obtained by using the same reconstruction parameters of the trabecular bone specimens. The cross section images of these specimens were obtained using the grey level scale, with white set to 0 value and black set to 255 value. The densitometric calibration protocol was divided in two steps: 1) Hounsfield Unit (HU) densitometric calibration by using water phantom and 2) bone mineral density (BMD) calibration by using densitometric phantoms. A density range calibration was necessary to change the grey level index scale in HU scale. It was performed by using all the reconstructed slices of distilled water phantom. After that the BMD calibration was performed. A stack of 481 and 466 reconstructed slices of 250mg/cm and 750mg/cm densitometric phantoms respectively were used to perform the BMD calibration. The same round region of interest (ROI), diameter 341pixel, was used during the application of the densitometric calibration protocol and its validation. According to the densitometric calibration protocol, the BMD was obtained by averaging the trabecular bone and the marrow density because the volumetric density of trabecular framework is generally under estimated due to partial volume effect. The BMD of trabecular bone specimens was associated with ρASH to validate the calibration protocol.