Comparing Gaussian and Bessel-Gauss beams for translating ultrafast laser ablation towards soft tissue surgery

Abstract

The goal of this research was to further improve existing ultrafast laser surgery techniques. To do so, different beam shapes (Bessel-Gauss and Gaussian) were compared for performing ultrashort picosecond pulsed surgery on various soft biological tissues, with the goal of minimising collateral thermal damage. Initially, theoretical modelling was performed using OpticStudio to test axicons of various conical angles. A 20° axicon was selected, but unfortunately early tests on murine intestinal tissue indicated a lack of sufficient intensity to achieve plasma-mediated ablation of the tissue with the 6ps input pulses of 85 µJ energy. Subsequently, a reimaged setup was designed in OpticStudio to demagnify the beam by a factor of 1.4x. The ability of this demagnified Bessel-Gauss beam to perform plasma-mediated ablation of murine intestinal tissue was confirmed through histological analysis. Another setup was also designed to produce a Gaussian beam of equivalent spot size. These beams were then tested on porcine intestinal tissue using lower pulse repetition rates of 1, 2 and 3 kHz, with optimal ablation and thermal damage margins of less than 20 µm (confirmed through histological analysis) being achieved with the Bessel-Gauss beam for spatial pulse overlaps of 70%, while for the Gaussian beam the prominence of cavitation bubble formation at both 2 and 3 kHz inhibited the respective ablation processes at this same spatial pulse overlap. As the numbers of passes were increased, the Bessel-Gauss beam also showed a trend of increased ablation depths. This was attributed to its large depth of focus of over 1 mm, compared to the theoretical 48 µm depth of focus for the Gaussian beam. After characterisation of fixated, non-ablated porcine intestine sample surfaces to quantify the inhomogeneity, another set of ablation trials was performed at higher pulse repetition rates (5, 10 and 20 kHz) to test more clinically viable processes. For the Bessel-Gauss beam, spatial pulse overlaps of up to around 50% at 5, 10 and 20 kHz offered excellent thermal confinement (with damage margins of < 30 µm, < 50 µm and < 25 µm respectively) and shape control, but at 70% and greater pulse overlaps the ablated feature became hard to control despite good thermal confinement (< 40 µm). The Gaussian beam, while having the advantage of achieving plasma formation at lower input pulse energies, was again found to be more prone to undesirable cavitation effects. Cavitation bubbles were observed in the histology images for spatial pulse overlaps as low as 15% for 5 kHz and 30% for both 10 and 20 kHz. From the histology images it is clear to see that these effects became more pronounced as the pulse repetition rate was increased. Conversely, the more consistent spot size of the Bessel-Gauss beam across its longer focal depth resulted in a higher tolerance to cavitation bubble formation. This was also demonstrated by high-speed videos of the beams being scanned across porcine skin samples. This could be significant as it may allow for higher ablation rates. In addition, it could ease the design constraint of the maximum speed at which the beam can be scanned at the distal end of an endoscopic device. Despite this, both beams were able to achieve distinct ablation with high thermal confinement for certain parameters. This work further highlights fibre-delivered ultrashort laser pulses as a promising alternative to existing endoscopic tumour resection techniques, which carry a higher risk of bowel perforation.