Postoperative Iris Biometric Identification After Cataract Surgery: A Prospective Comparative Cohort Study of Associated Surgical and Intraoperative Factors

Background and Objectives: Iris-based biometric identification is among the most accurate personal recognition technologies in routine use today. Cataract remains the leading cause of blindness worldwide, and surgical removal of the opacified lens is the most frequently performed ophthalmic operation globally, inevitably involving mechanical and thermal manipulation of the anterior segment, where the iris resides. The aims of this study were: (1) to compare the effect of ultrasound phacoemulsification (PHACO) and conventional extracapsular cataract extraction (ECCE) on iris-based identification; (2) to determine whether the vital dye trypan blue, used during phacoemulsification of mature cataracts, independently impairs biometric accuracy; and (3) to quantify associations between key intraoperative parameters and post-operative recognition performance. Materials and Methods: This prospective, non-randomised comparative cohort study enrolled 120 participants assigned to four parallel cohorts (n = 30 each) (clinically allocated by surgical indication rather than randomised; group comparability with respect to age, cataract severity and case complexity was not guaranteed, and findings should therefore be interpreted as observational associations rather than causal effects) according to clinical indication: PHACO (uncomplicated phacoemulsification), PHACO-DYE (phacoemulsification with trypan blue 0.6 mg/mL for mature cataracts), ECCE (conventional extracapsular extraction for complex cataracts), and a healthy Control group. Standardised iris photographs were acquired on the day before the surgery and one month after surgery using a semi-professional digital camera under controlled lighting. Images were processed by a custom Java-based iris recognition platform that performed boundary detection, polar normalisation into 65 discrete iris sectors, and sector-by-sector comparison using a dual-threshold matching rule: sectors with 100% identity, or with a residual difference below 2%, were scored as direct matches; all others were scored as indirect, non-matching elements. The primary outcome was the percentage of correspondence (match %) between pre- and post-operative templates. Intraoperative parameters, cumulative dissipated energy (CDE), phaco time, and estimated fluid used, were recorded for all phacoemulsification cases. Results: All three surgical groups showed significantly lower match percentages than controls (median 91.0%, IQR 86–96; Mann–Whitney U test, all p < 0.001). Median match % was 85.0% (IQR 71–89) for PHACO, 78.5% (68–85) for PHACO-DYE, and 44.0% (29–61) for ECCE. ECCE differed significantly from both phacoemulsification groups (Mann–Whitney U test, p < 0.001 each); no significant difference was found between PHACO and PHACO-DYE (Mann–Whitney U test, p = 0.210, not statistically significant). Only CDE differed significantly between PHACO and PHACO-DYE (Mann–Whitney U test, p = 0.026, statistically significant). All three intraoperative parameters showed significant negative correlations with match % (Pearson correlation coefficient), with estimated fluid used showing the strongest association (PHACO: r = −0.88; PHACO-DYE: r = −0.71; both p < 0.001). Conclusions: In this non-randomised cohort, surgical group was associated with postoperative match reduction, with ECCE showing markedly greater reduction than phacoemulsification; however, residual confounding by cataract severity, case complexity and baseline age differences limits causal interpretation. No significant unadjusted difference in match % was found between PHACO and PHACO-DYE, indicating that this study could not demonstrate an independent effect of trypan blue, though it also cannot exclude one, since dye use was itself linked to more complex, mature cataracts. Estimated intraoperative fluid use showed the strongest unadjusted negative correlation with match %; whether it is an independent predictor requires multivariable analysis that this study, as designed, was not powered to perform. These preliminary, hypothesis-generating findings support considering postoperative biometric re-enrolment after cataract surgery and further, adequately powered and adjusted studies before firm clinical or algorithmic recommendations can be made.