We have previously applied this simulation model to ocular injury by an airbag impact in several studies [20-22]. In these reports, a surgical procedure was added in the simulation eye model, such as radial keratotomy , photorefractive keratectomy  or transscleral fixation of a posterior chamber intraocular lens . In contrast, this report examined a series of airbag-induced injury in our simulation studies performed on an intact eye, which comprises most cases suffering from airbag deployment in clinical situations. In this study, we reported that considerable damage, such as corneal or scleral laceration, was observed especially at higher impact velocities, such as 50 or 60 m/s, at any axial length. This seems reasonable because greater physical energy is derived from a higher airbag impact velocity. It has been reported that airbag-related eye injuries occurred very rarely in car accidents in cases where the occupant survived and the restraint system was appropriately used , and airbags had no significant effect to increase eye injuries in studies on motor vehicle collisions . Hwang et al. recommended that when using airbags, seat belts should be used together, because seat belts were effective to decrease eye injuries in motor vehicle collisions . From these studies, serious ocular injuries induced by airbag deployment in car accidents are expected to be limited to cases such as improper use of a restraint system or passengers with short stature.
A limited number of studies have evaluated the relationship between axial length and severity of blunt ocular trauma [33-35]. Rau et al. reported in an experimental study using human cadaveric eyes that the force at corneal rupture was not associated with axial length . However, Hashemi et al reported that axial length was significantly longer in cases with a history of trauma in a cross-sectional study of a clinical database of ocular trauma cases . Even considering the difference in design of these studies, their results were contradictory. Therefore, we have carried out several FEAs on blunt ocular trauma impact by an airbag or airsoft gun in eyes of various axial lengths using a simulation eye model we have established [22, 24]. While hyperopic eyes are most susceptible to deformation by an airsoft gun impact compared with other axial length eye models , eyes with long axial length (myopic) experienced greater deformity upon airbag impact . The reason for this discordance is unclear, but it might be due to difference in outcome measures between these studies, because the latter study focused on the upper limit of the impact force exceeding the tensile force of 10-0 polypropylene that caused breakage of sutures ; in contrast, the former report mainly evaluated strain strength threshold of the cornea and sclera .
As indicated above, the results of the present study are consistent with our previous study of an FEA model to evaluate airsoft gun impact on an eye and the deformation rate of eyes of various axial lengths at various velocities . It is also interesting that deformity of posterior segments including the vitreous body and retina was less than that of the anterior chamber, especially in the hyperopic eye, in sagittal images (Figure 4); however; the degree of retinal or posterior scleral deformity was most apparent in the hyperopic eye (Figure 4). Our previous report of airsoft gun impact in which deformation was most evident in the anterior segment, while deformation of the posterior segment was less, also supports the results of this study. The degree of tissue strain or area of deformation was greatest in the shorter axial eye followed by the normal eye and the longer axial eye compared at the same airbag impact velocity in the simulation in this study; however, the difference seems not so evident. The impact velocities set in this simulation study were within those of current vehicles and were chosen linearly; thus, the results obtained did not show apparent differences. In clinical situations, other factors, such as head movement, might add further irregular physical impact on the eyeball; however, these were not included in the variables in this simulation. These could be the reasons for the lack of an evident tendency of ocular deformation in this study. Considering the rare occurrence of airbag-related eye injuries, if the restraint system was used appropriately , our results seem consistent with the rarity of serious ocular blunt injury caused by airbag impact. However, the present result that the cornea and sclera might suffer considerable strain that might result in laceration of the eyeball at higher impact velocities (50 or 60 m/s) emphasized the necessity of wearing a seatbelt or eye protection, such as glasses, to prevent serious ocular injury especially in high-risk populations, such as short-stature persons or small children in the passenger seat.
As for decrease rate of axial length, the present result was not inconsistent with our previous study of airsoft gun impact simulation . The highest decrease rate was observed in the hyperopic eye in both studies; however, it was followed by the myopic eye in this study (Table 1), while the emmetropic eye was the second highest in the airsoft gun simulation study . It is difficult to determine the exact reason; however, the larger area of compression and higher energy with airbags caused a more considerable effect on the myopic eye, in which ocular wall thickness became thinner according to the extension of the volume of the eyeball, compared with the normal length eye with a thicker ocular wall.
Despite our careful calculation in the simulation model, there are still limitations to our study. First, in order to reduce the time for computer calculation, the impacting object was placed almost adjacent to the surface of the eyeball for impact simulation. Therefore, further study on the distance of the deploying airbag from the driver in regard to the change in velocity would be necessary to determine the impact that is likely to occur in reality. Secondly, the vitreous model was a solid mass with physiological intraocular pressure of 20 mmHg . Factors associated with aging such as vitreous viscosity and other physical characteristics were not considered as variables in this simulation. However, we believe that our goal to assess the mechanical properties of eyes with different axial lengths impacted by an airbag was achieved by demonstrating that corneoscleral laceration might occur at higher impact velocity in eyes with shorter axial length, being more susceptible to impact injuries.