Kumar Sonali Vinay, Kumar Vinay, Kumar Natasha Vinay, Sati Alok, Mishra Sanjay Kumar
Command Hospital Eastern Command, Kolkata, West Bengal, India.
JIS School of Medical Science and Research, Howrah, West Bengal, India.
Rom J Ophthalmol. 2025 Apr-Jun;69(2):184-189. doi: 10.22336/rjo.2025.30.
This study investigates the impact of cataract surgery on the reliability and accuracy of iris recognition systems used for biometric identification.
We carried out a prospective observational study on patients undergoing cataract surgery via phacoemulsification. The study comprised 100 participants who underwent cataract surgery. We recorded preoperative iris scans using a standard biometric device (iris camera) and rescanned the same patients after surgery. Matching scores before and after surgery were analyzed to detect discrepancies. We examined the Hamming distance to evaluate changes in the iris pattern.
The mean age of the study population was 64 ± 6.2 years (range, 52-74 years). The study population comprised 62 males and 38 females. Out of the total cases, 72 involved the right eye and 28 involved the left eye. The majority of patients presented within the current study presented with visual acuity between 6/18 and 3/60. We performed phacoemulsification and implanted foldable intraocular lenses in all cases. The study found no notable changes in iris pattern matching accuracy following uncomplicated cataract surgery. Postoperative images were matched with preoperative images in 95 patients, demonstrating that the surgery did not affect biometric reliability. The mean Hamming distance before and after surgery remained within the acceptable threshold for authentication, with 99% of cases successfully matched.
The study supports the robustness of iris-based biometric systems under routine ophthalmic surgical conditions. This system's resilience to variations caused by cataract surgery demonstrates its robustness in practical sessions.
Cataract surgery, when performed without complications, does not impair the accuracy of iris-based biometric identification. These findings underscore the feasibility of using iris recognition systems in healthcare and security, even among patients undergoing ocular procedures.
本研究调查白内障手术对用于生物识别的虹膜识别系统的可靠性和准确性的影响。
我们对接受超声乳化白内障手术的患者进行了一项前瞻性观察研究。该研究包括100名接受白内障手术的参与者。我们使用标准生物识别设备（虹膜相机）记录术前虹膜扫描，并在术后对同一患者进行重新扫描。分析手术前后的匹配分数以检测差异。我们检查汉明距离以评估虹膜模式的变化。
研究人群的平均年龄为64±6.2岁（范围52 - 74岁）。研究人群包括62名男性和38名女性。在所有病例中，72例涉及右眼，28例涉及左眼。本研究中的大多数患者视力在6/18至3/60之间。所有病例均进行了超声乳化并植入了可折叠人工晶状体。研究发现，在无并发症的白内障手术后，虹膜模式匹配准确性没有明显变化。95例患者的术后图像与术前图像匹配，表明手术未影响生物识别可靠性。手术前后的平均汉明距离保持在可接受的认证阈值内，99%的病例成功匹配。
该研究支持基于虹膜的生物识别系统在常规眼科手术条件下的稳健性。该系统对白内障手术引起的变化的恢复能力证明了其在实际应用中的稳健性。
在无并发症的情况下进行白内障手术不会损害基于虹膜的生物识别的准确性。这些发现强调了在医疗保健和安全领域使用虹膜识别系统的可行性，即使是在接受眼部手术的患者中。
