OBJECTIVES: To assess the effect of acquisition duration (T(acq)) and pre-enhancement set points (T₁) on computer tomography perfusion (CT(p)) parameter values in primary and metastatic tumors in the lung. MATERIALS AND METHODS: 24 lung CT(p) datasets (10 primary; 14 metastatic), acquired using a two phase protocol spanning 125 s, in 12 patients with lung tumors, were analyzed by deconvolution modeling to yield tumor blood flow (BF), blood volume (BV), mean transit time (MTT), and permeability (PS) values. CT(p) analyses were undertaken for the reference dataset (i.e., T₁=t₀) with varying T(acq) from 12 to 125 s. This was repeated for shifts in T₁ (±0.5 s, ±1.0 s, ±2.0 s relative to the reference at t₀). Resultant CTp values were plotted against T(acq); values at 30 s, 50 s, 65 s and 125 s were compared using linear mixed model. RESULTS: All CT(p) parameter values were noticeably influenced by T(acq), with generally less marked changes beyond 50 s, and with no difference in behavior between primary and secondary tumors. Apart from BV, which attained a plateau at approximately 50s, the other three CT(p) parameters did not reach steady-state values within the available 125 s of data, with values at 30 s, 50 s and 65 s significantly different from 125 s (p<0.004). Shifts in T₁ also affected the CT(p) parameters values, with positive shifts having greater impact on CT(p) values than negative shifts. CONCLUSION: CT(p) parameter values derived from deconvolution modeling can be markedly affected by T(acq), and pre-enhancement set-points. 50 s acquisition may be adequate for BV, but longer than 125 s is probably required for reliable characterization of the other three CT(p) parameters.
OBJECTIVES: To assess the effect of acquisition duration (T(acq)) and pre-enhancement set points (T₁) on computer tomography perfusion (CT(p)) parameter values in primary and metastatic tumors in the lung. MATERIALS AND METHODS: 24 lung CT(p) datasets (10 primary; 14 metastatic), acquired using a two phase protocol spanning 125 s, in 12 patients with lung tumors, were analyzed by deconvolution modeling to yield tumor blood flow (BF), blood volume (BV), mean transit time (MTT), and permeability (PS) values. CT(p) analyses were undertaken for the reference dataset (i.e., T₁=t₀) with varying T(acq) from 12 to 125 s. This was repeated for shifts in T₁ (±0.5 s, ±1.0 s, ±2.0 s relative to the reference at t₀). Resultant CTp values were plotted against T(acq); values at 30 s, 50 s, 65 s and 125 s were compared using linear mixed model. RESULTS: All CT(p) parameter values were noticeably influenced by T(acq), with generally less marked changes beyond 50 s, and with no difference in behavior between primary and secondary tumors. Apart from BV, which attained a plateau at approximately 50s, the other three CT(p) parameters did not reach steady-state values within the available 125 s of data, with values at 30 s, 50 s and 65 s significantly different from 125 s (p<0.004). Shifts in T₁ also affected the CT(p) parameters values, with positive shifts having greater impact on CT(p) values than negative shifts. CONCLUSION: CT(p) parameter values derived from deconvolution modeling can be markedly affected by T(acq), and pre-enhancement set-points. 50 s acquisition may be adequate for BV, but longer than 125 s is probably required for reliable characterization of the other three CT(p) parameters.
Authors: Jinzhong Yang; Lifei Zhang; Xenia J Fave; David V Fried; Francesco C Stingo; Chaan S Ng; Laurence E Court Journal: Comput Med Imaging Graph Date: 2015-12-14 Impact factor: 4.790