Hydrodynamic Modeling of Pulse Structure Effects during Free Electron Laser Ablation of Tissue

(Wrk-P-3-05)

Stephen R. Uhlhorn, Richard A. London, Anthony J. Makarewicz, Jonathan M. Gilligan, E. Duco Jansen  

Departments of Biomedical Engineering (SRU,EDJ) and Physics (JMG), Vanderbilt University, Nashville, TN 37235; and Lawrence Livermore National Laboratory (SRU, RAL, AJM), Livermore, CA, 94551

The Vanderbilt free electron laser (FEL) provides a continuously tunable (l = 2-10 $\mu$m) source of pulsed IR radiation. Previous research has focused on exploiting the wavelength tunability of the FEL to target protein absorption bands in the mid-IR for efficient tissue ablation with minimal collateral damage. However, the temporal pulse structure of the laser output is unlike those of conventional lasers. Despite extensive studies of the effect of wavelength on ablative properties of primarily neural tissue, leading to the first human clinical application of the FEL as a surgical laser, the effect of the pulse structure on the ablation process has not been investigated. An investigation as to the role of the FEL pulse structure in tissue ablation is performed using theoretical modeling and experimental measurements. A hydrodynamic code, known as LATIS3D, is used to model the ablation of tissue using the FEL. Modeling was performed by comparing two scenarios of temporal energy deposition. First, a sample is irradiated with a laser pulse train similar to that of the FEL (1 ps micropulse, 333 ps micropulse separation, 100| $\mu$m spotsize, wavelength = 6.45 $\mu$m, depth of penetration = 10 $\mu$m, energy/micropulse = 0.3 $\mu$J, total energy deposited = 50 $\mu$J). For practical and computational reasons we limited the duration of the pulse train to 150 ns. In the second set of simulations, identical parameters were applied except the energy of 50 $\mu$J was deposited as a time averaged single pulse with a duration of 150 ns. The modeling results suggest that the pulse train irradiation creates large pressure gradients in the sample on the order of 300 bar /  $\mu$m, coinciding with each individual micropulse, whereas the time averaged pulse did not. The macroscopic pressure transient was similar for both scenarios. Results of another output parameter of the simulation, namely surface displacement are compared to experiments using interferometric surface monitoring of surface displacement.