[Consistency between digital navigation and slit-lamp assisted corneal marking].

<b>Objective:</b> To investigate the consistency between automatic corneal marking using a surgical navigation system and manual corneal marking assisted by a slit-lamp microscope. <b>Methods:</b> This was a retrospective case series study. Patient data from the Ophthalmology Outpatient Department of Shanghai Sixth People's Hospital, Shanghai Jiao Tong University School of Medicine, from June 2021 to July 2023 were continuously collected. All patients underwent manual corneal marking assisted by a slit-lamp microscope, followed by automatic marking using a surgical navigation system. Patients were divided into two groups: the manual marking group and the automatic marking group. The manual marking group was further divided into three subgroups based on the marking measurement methods: manual marking (upper to lower edge), manual marking (lower to upper edge), and manual marking (midpoint). The horizontal meridian angle and the deviation of the central axis were calculated for the four marking methods using the image processing software. Statistical differences among the measurement methods were compared, and the consistency was evaluated using the Bland-Altman plots and intraclass correlation coefficient (ICC). <b>Results:</b> A total of 143 cataract patients (201 eyes) were included in the study, with an average age of (60.27±19.13) years, including 58 males and 85 females. The absolute mean deviations of the horizontal meridian angle for manual marking (upper to lower edge), manual marking (lower to upper edge), manual marking (midpoint), and automatic marking were 3.61°, 4.76°, 3.20°, and 2.45°, respectively, with statistically significant differences among them (<i>P</i><0.001). The differences between automatic marking and the three manual marking methods were also statistically significant (<i>P</i><0.05). The mean deviations of the central axis were 2.35, 2.35, 2.24, and 0.40 mm, respectively, with statistically significant differences among them (<i>P</i><0.001). The deviation of the automatic marking method was significantly lower than that of the manual marking methods (<i>P</i><0.001), while no significant difference was observed between the three edge-based manual marking methods (<i>P</i>>0.05). The ICC for the horizontal meridian angle between manual marking (midpoint) and automatic marking was 0.88, while the ICC for the central axis deviation was -0.04, indicating good consistency in the horizontal meridian angle between manual marking (midpoint) and automatic marking, but poor consistency in the central axis deviation. <b>Conclusions:</b> The horizontal meridian angle values of automatic corneal marking using a surgical navigation system and manual corneal marking (midpoint) assisted by a slit-lamp microscope showed good consistency. However, digital navigation automatic marking demonstrated significantly better control in the centralization of the radial offset, providing a robust basis for accurate axis positioning during toric intraocular lens implantation.