Prediction of CW chord as a measure for the eye's orientation axis after cataract surgery from preoperative IOLMaster 700 measurement data.

BACKGROUND The angles alpha and kappa are widely discussed for centring refractive procedures, but they cannot be determined with ophthalmic instruments. The purpose of this study is to investigate the Chang-Waring chord (position of the Purkinje reflex PI relative to the corneal centre) derived from an optical biometer before and after cataract surgery and to study the changes resulting from cataract surgery. METHODS The analysis was based on a large dataset of 1587 complete sets of preoperative and postoperative IOMaster 700 biometry measurements from two clinical centres, each containing: valid data for pupil and corneal centre position, the position of the Purkinje reflex PI originated from a coaxial fixation target, keratometry (K), axial length (AL), anterior chamber depth (ACD), lens thickness (LT), central corneal thickness CCT, and horizontal corneal diameter W2W. The Chang-Waring chord CW was derived from pupil centre and Purkinje reflex PI analysed preoperatively and postoperatively, and a multilinear regression model together with a feedforward neural network algorithm was set up to predict postoperative CW chord from preoperative CW chord, K and biometric distances of the eye. RESULTS The Y component of CW chord shows a slight shift in the inferior direction in both left and right eyes, before and after cataract surgery. The X component shows some shift in the temporal direction, which is more pronounced preoperatively and slightly reduced postoperatively but with a larger variation. The change in CW chord from preoperative to postoperative shows a slight shift in the superior and nasal directions. Our algorithms for prediction of postoperative CW chord using preoperative CW chord, keratometry and biometry as input data performed with a multilinear regression and a feedforward neural network approach were able to reduce the variance, but could not properly predict the postoperative CW chord X and Y components. CONCLUSION The CW chord as the position of the Purkinje reflex PI with respect to the pupil centre can be directly measured with any biometer, topographer or tomographer with a coaxial fixation light. The mean Y component does not differ between right and left eyes or preoperatively and postoperatively, but the mean temporal shift of the X component preoperatively is slightly reduced postoperatively, but with a larger scatter of the values.