Influence of beam hardening correction (BHC) on Hounsfield unit accuracy: Limitations of using Teflon as a bone-like...

To compare Hounsfield Unit (HU) measurements in two phantoms simulating various structures with different electronic densities and evaluate the influence of beam hardening correction (BHC) on HU variations.

HU values were assessed using the Canon® Aquilion Lightning scanner under an abdominal exposure protocol of 120 kVp. The study analyzed 16 convolution filters (CFs), eight with BHC and eight without, across two phantoms: Catphan® 504 and CIRS 062M. Measurements were conducted using ImageJ software.

The application of BHC did not significantly alter HU values in low-electron-density materials such as Lung (inhale), Liver, and Adipose, with mean variations of -3.55 HU for Lung and ＋2.02 HU for Liver in the CIRS phantom. However, high-electron-density materials, including Teflon®, Bone 200, and Bone 800, exhibited significant HU variations (p ≤ 0.002), with HU differences ranging from -35.0 HU to -80.3 HU. The largest variation was observed in Teflon® (-80.3 HU), whereas Bone 200 and Bone 800 showed changes of -18.3 HU and -11.3 HU, respectively. Additionally, materials located in the central region of the CIRS phantom demonstrated greater HU variation, with the Liver showing a HU difference of 14.9 HU. BHC significantly influences HU values in high-electron-density materials, with the extent of this effect depending on the physical and chemical properties of the materials. Teflon® exhibited the largest HU variation, reinforcing its inadequacy for mimicking dense bone, while Delrin® displayed smaller but still notable variations. Materials from the CIRS 062M phantom showed variations within clinically acceptable limits, highlighting the importance of selecting more representative materials for accurate tissue simulation. These findings emphasize the necessity of using phantoms with materials such as those in CIRS 062M to ensure consistent and reliable HU measurements in CT imaging.