Investigation Short-Term Stability Of Iols With Different Haptic Designs Using Cnn-Integrated Ss-Oct

Utilizing Convolutional Neural Networks (CNN) in combination with Swept-Source Optical Coherence Tomography (SS-OCT) to investigate the spatial positional differences and influencing factors of intraocular lenses (IOL) with plate-haptic and C-loop haptic designs following cataract surgery.

Shanxi Eye Hospital Affiliated to Shanxi Medical University, Shanxi, China.

The participants were divided into two groups: Group A with the ATTORBI 709M plate-haptic design (44 eyes), while Group B with the PY-60AD C-loop design (44 eyes). SS-OCT examinations were conducted both prior to and one month following the cataract surgery. A CNN algorithm was employed to construct an analytical model that successfully generated three-dimensional spatial data regarding IOL tilt and decentration using the visual axis as the reference. Univariate and multivariate linear regression analyses were conducted to identify factors influencing postoperative IOL tilt and decentration within each group. An independent-sample t-test was utilized to compare these postoperative characteristics between the two groups.

This study included 79 patients (88 eyes). Prior to surgery, there were no statistically significant differences in ocular parameters between the two groups (P>0.05). A significant difference in IOL decentration was noted between the two groups (Group A & Group B: 0.15±0.08 mm & 0.41±0.28 mm, P<0.001), while the difference in tilt was not significant (Group A & Group B: 4.59±1.34° & 4.64±1.91°, P=0.903). Both groups showed a predominant tilt of the crystalline lens and IOL towards the inferotemporal directions, whereas decentration did not display any distinct distribution pattern. Multivariate linear regression confirmed that IOL tilt in both groups was correlated to crystalline lens tilt (P<0.002). 

The plate-haptic design IOL demonstrates superior short-term stability after cataract surgery compared to the C-loop design.