Creating a lenticule with good surface quality is essential in SMILE, because it has been proved that lenticule quality is significantly correlated with the incidence of vision-threatening complications and refractive outcomes. [9-11]Therefore, it is important to guarantee lenticule surface quality in SMILE for hyperopia correction. In the current study, we evaluated lenticule surface characteristics after SMILE for hyperopia correction in rabbits for the first time.
Tissue bridges were distributed evenly in most areas of the lenticules. It has been reported that tissue bridges were the main entity for surface irregularities on the corneal lenticule surface.It was described as residual fibers between the interfaces after completion of laser cuts. Laser cutting, as well as surgical manipulation, could affect the arrangement of tissue bridges. In the current study, although the laser settings were exactly the same as those used in the previous study evaluating human lenticules, tissue bridges in rabbit lenticules were longer and looser compared to those of human lenticules. There are two potential reasons for this. One is that the structure of corneal collagen fibers in rabbits may differ from that of humans. Another possibility is that the rabbits had bad coordination during the operation and required more surgical dissection, which may have affect the surface quality and cause the roughness.
The lenticules exhibited comparable surface characteristics in both of the groups in the study, and the central and periphery regions of the posterior surfaces of the hyperopia lenticules exhibited similar arranged tissue bridges. With regard to the treatment of hyperopia, it has been reported that lenticule shape was thicker in the periphery and thinner in the center. [13,14]This may increase the difficulty of surgical manipulation and result in surface roughness. To reduce operative difficulty, the minimal central thickness of hyperopic lenticule was limited to 25 µm in the current study. Also, the surgeon had extensive experience in the performance of SMILE and performed the procedure gently, and thus did not increase the surface roughness of hyperopia lenticules. Notably however, among the 3 cases that exhibited partially rough surfaces, in 2 cases it was related to the posterior surface and in only 1 case it was related to the anterior surface The results suggested that dissection between the lenticule-stromal bed interface could be more difficult than dissection at the cap-lenticule interface: the anterior surface can be dissected under the cap smoothly, but for the posterior surface, dissection was done more slowly and involved repetition action because the lenticule was thin and unfixed. This emphasizes the importance of good cooperation and careful surgical manipulation during the procedure, for both hyperopia and myopia correction.
Visible cavitation holes were absent in all lenticules except 1 in the myopia group that exhibited a few cavitation holes on the anterior surface. Cavitation holes are gas bubbles formed during the vaporization of the corneal tissue, and they affect scanning quality. The scanning quality of the anterior surfaces of the lenticules was comparable in the two groups. There was no significant difference in scanning quality of the posterior surface either, despite the fact that the lenticule-stromal interface was concave in the hyperopia group but convex in the myopia group. The working principle of the femtosecond laser may explain this finding. The femtosecond laser is a near-infrared laser that produces ultra-short pulses of light. In the non-thermal ablation process that is achieved via corneal photodisruption, a plasma state develops with optical breakdown, and some cavitation gas bubbles are formed. A series of bubbles is created resulting in separating of the corneal tissue at a precise depth. On the basis of the mechanism of femtosecond laser scanning, the efficacy of laser scanning is not affected by the shape of the interface, thus the two types of lenticule are indistinguishable with regard to the scan surface.
Previous studies had investigated the corneal surface characteristics in myopia correction during SMILE, and reported that pulse energy and laser frequency are two foremost parameters influencing scanning quality.  Heichel  firstly reported scoring of lenticule surface quality in porcine corneas using the original VisuMax femtosecond laser system with a repetition rate of 200 kHz and a pulse energy of 185 nJ. Although lenticules of predictable surface quality were created in the procedure, the results suggested that laser settings should be improved. Kunert  evaluated lenticule surface characteristics with a fixed repetition rate of 200 kHz and different energy levels (150, 180, and 195 nJ). The highest surface regularity score was achieved using the lowest pulse energy, presenting that lower pulse energy facilitates a smoother cut surface. Furthermore, smoother lenticule surfaces were reported observed using a higher frequency laser.[18,8]
Based on the aforementioned investigations, a new generation VisuMax laser with settings of a 500 kHz repetition rate and a pulse energy of 130 nJ was investigated. Researchers studied the scan quality of corneal lenticules in the context of myopia treatment using the new laser system, and found that both sides of the lenticules exhibited smooth surfaces.  Therefore, in the current study, the exact same laser settings were used to perform hyperopia correction in SMILE, and enabled homogeneous cutting.
The current study had some limitations, the foremost being the small number of specimens involved. Also, while precautions were taken during preparation for imaging, scratches and grooves may have been generated. Lastly, the refractive outcomes could not be evaluated in the animal model.