Latent density discrepancies in commercial lung‐equivalent inserts and their clinical dosimetric impact
Minoru Nakao, Shuichi Ozawa, Hideharu Miura, Masahiro Hayata, Kosaku Habara, Masahiro KenjoAbstract
Background
Accurate heterogeneity correction in high‐precision radiotherapy relies on precise computed tomography (CT) number‐to‐density conversion via the Hounsfield unit look‐up table (HLUT). While physical properties of tissue‐equivalent materials are generally assumed consistent with manufacturer specifications, an independent audit identified clinically significant density discrepancies in commercially available lung‐equivalent phantom inserts.
Purpose
This study evaluates the physical properties of nonconforming lung inserts through mass measurements and stoichiometric analysis, and assesses the clinical dosimetric impact of the associated density discrepancies.
Methods
Five lung‐inhale inserts manufactured in 2010, 2015, and 2024 (10A, 10B, 15A, 15B, and 24A) were analyzed. Mass and physical dimensions were measured in triplicate using a precision balance (1 mg resolution) and vernier calipers. Stoichiometric analysis was conducted using reference materials to evaluate the tissue‐equivalence of the inserts and quantify deviations from the theoretical baseline. A nonconforming table and a conforming reference table (RT) were established, derived from inserts 10A and 24A, respectively. For clinical impact assessment, volumetric modulated arc therapy (VMAT) plans for three clinical cases involving centrally located lung tumors (utilizing both inspiration breath‐hold (IBH) and free‐breathing) were optimized for stereotactic body radiotherapy (SBRT) and recalculated with the RT using the Acuros XB algorithm. Differences in gross tumor volume (GTV) mean dose and planning target volume (PTV) D95% were evaluated to quantify the dosimetric consequences.
Results
The 2010 inserts (10A and 10B) exhibited a 17.1% mass reduction and lower CT numbers compared to the reference 24A insert. Dimensional variations were negligible (≤ 0.2 mm) across all samples. Clinical recalculation revealed maximum dose reductions of 2.1% for the GTV mean dose and 3.0% for the PTV D95% in the worst‐case scenario. These errors exceed the 2% clinical tolerance, propagated by HLUT interpolation across the low‐density range.
Conclusions
Substantial inter‐lot density variations in commercial calibration phantoms can lead to dosimetric errors that exceed established clinical limits, particularly for centrally located tumors treated with IBH. Medical physicists must not implicitly rely on nominal manufacturer values; independent audits and initial mass screening at acceptance are highly recommended for maintaining dose calculation accuracy.