$$\rightleftharpoonup{xx}$$
$$\longleftharp{xx}$$,
$$\longrightharp{xx}$$,
The small incision lenticule extraction (SMILE) is an advanced type of corneal refractive surgery, and has a number of benefits over conventional laser in situ keratomileusis (LASIK)1,2,3,4. Now, SMILE has become mainstream practice in the correction of myopia and astigmatism5,6. Nevertheless, the SMILE surgery is prone to causing under-correction and regression in the correction of patients with moderate to high myopia and astigmatism7,8. Multiple factors contribute to the variability in astigmatic correction after SMILE, including the magnitude and type of preoperative astigmatism, surgical centration, Kappa angle, and ocular rotation9,10,11. Among these, ocular rotation (positional cyclotorsion resulting from changes in patient posture) is identified by Chow et al. as one of the most crucial variables influencing the accuracy of astigmatic correction12,13. This is because the femtosecond laser system does not have an integrated cyclotorsion tracking14,15.
In order to obtain precise quantitative data on the magnitude of ocular rotation, a goal that was previously difficult to achieve. Other researchers proposed an artificial intelligence (AI)-powered system for estimating the rotation angle of the eyeball, which requires accurate segmentation of the optic disc and macula and calculates the rotation angle based on these features16,17. The approach, however, is operationally difficult and not suitable for corneal refractive surgeries such as SMILE. Therefore, we developed a standardized and objective method - the double corneal-scleral marking. The quantifiable data generated by this method hold fundamental value and constitute a key tool for systematically studying the impact of ocular rotation in SMILE surgery. Based on this measurement, it is crucial to deepen research on postoperative astigmatism, construct a surgical accuracy prediction model based on multi-center and big data, and build an ocular rotation amplitude prediction model based on ocular features and AI.
In summary, this simple, repeatable, and easily integrated into clinical practice solution, by reliably capturing the intraoperative rotation indicators, compensates for the shortcomings of existing research methods and provides a solid data-driven research foundation for ultimately achieving better accuracy and predictability in managing SMILE surgery-induced astigmatism and optimizing the clinical treatment pathway.