Tilt and decentration of intraocular lenses in vivo from Purkinje and Scheimpflug imaging. Validation study.

PURPOSE

To measure tilt and decentration of intraocular lenses (IOLs) with Scheimpflug and Purkinje imaging systems in physical model eyes with known amounts of tilt and decentration and patients.

SETTING

Instituto de Optica Daza de Valdés, Consejo Superior de Investigaciones Científicas, Madrid, Spain.

METHODS

Measurements of IOL tilt and decentration were obtained using a commercial Scheimpflug system (Pentacam, Oculus), custom algorithms, and a custom-built Purkinje imaging apparatus. Twenty-five Scheimpflug images of the anterior segment of the eye were obtained at different meridians. Custom algorithms were used to process the images (correction of geometrical distortion, edge detection, and curve fittings). Intraocular lens tilt and decentration were estimated by fitting sinusoidal functions to the projections of the pupillary axis and IOL axis in each image. The Purkinje imaging system captures pupil images showing reflections of light from the anterior corneal surface and anterior and posterior lens surfaces. Custom algorithms were used to detect the Purkinje image locations and estimate IOL tilt and decentration based on a linear system equation and computer eye models with individual biometry. Both methods were validated with a physical model eye in which IOL tilt and decentration can be set nominally. Twenty-one eyes of 12 patients with IOLs were measured with both systems.

RESULTS

Measurements of the physical model eye showed an absolute discrepancy between nominal and measured values of 0.279 degree (Purkinje) and 0.243 degree (Scheimpflug) for tilt and 0.094 mm (Purkinje) and 0.228 mm (Scheimpflug) for decentration. In patients, the mean tilt was less than 2.6 degrees and the mean decentration less than 0.4 mm. Both techniques showed mirror symmetry between right eyes and left eyes for tilt around the vertical axis and for decentration in the horizontal axis.

CONCLUSIONS

Both systems showed high reproducibility. Validation experiments on physical model eyes showed slightly higher accuracy with the Purkinje method than the Scheimpflug imaging method. Horizontal measurements of patients with both techniques were highly correlated. The IOLs tended to be tilted and decentered nasally in most patients.