Moderator
Seth M. Pantanelli, MD, MS, ABO
Panelists
Li Wang, MD, PhD
Nir Sorkin, MD
Viewing Papers
Expand a paper title to the right to view the paper abstract and authors. Use the video link to jump to that poster in the session.
Presenting Author
Matteo Piovella, MD
Co-Authors
Barbara Kusa MD
Purpose
To evaluate three physical consecutive treatments, Blephex for Lid Scrub(BlephEx, Brentwood TN) LipiFlow (J&J - Santa Ana Ca) ILux (Alcon Fort Worth Texas) for the thermal pulsation treatment of Meibomian Gland Dysfunction(MGD) in vaporative Dry Eye to improve quality of the surface of the cornea. to get precise biometry IOLs power calculations
Methods
Since 2017, 378 patients (mean age 66.58 �11.55 years) were treated for MGD. And Evaporative Dry Eye Patients received a LipiFlow treatment to remove obstructions and restore meibomian gland function. 123 of these patients received also Blephex treatment immediately before since September 2019. 65 patients since September 2023 adopted also ILux treatment to complete the MAPHRY protocol based on three physical treatments BlephEx,Lipiflow and Ilux.Adopting single treatment do not matches positive results. On the countrary three consecutive tratments in a single session provide best outcomes
Results
Postop quality of vision improved in all patients, and regular cornea surface provided more precise and stable biometry results. The adoption of blephex lipiflow and ILux treatments provided 97% of eyes inside the planned refractive postoperative outcome after cataract surgery.We experienced that in specific cases would be important to applay up to 5 times the treatments despite the high costs.
Conclusion
These treatments have a priority role in adopting implants to correct refractive defects and treat presbyopia in cataract patients. They improved precision in detecting the right presbyopic IOL power to avoid postoperative low quality of vision
Presenting Author
Aaron Ng, FRCOphth
Co-Authors
Tun Kuan Yeo FRCOphth
Purpose
To evaluate intraocular lens (IOL) power calculations in eyes with zonulysis where anterior chamber depth (ACD) and lens thickness (LT) may be non-physiological from lens subluxation.
Methods
Retrospective consecutive case series of 47 eyes that underwent cataract surgery with at least one capsular tension segment. Formulae compared include Barrett Universal II(BUII), EVO2.0(EVO), Hoffer Q, Holladay I and SRK/T. BUII and EVO were evaluated using all parameters(Full), axial length(AL) + keratometry(K), and AL + K + ACD. If the fellow eye (FE) was deemed to have normal physiologic parameters available, BUII and EVO was evaluated with FE:ACD+LT. The Eyetemis analysis tool was used to assess trueness (spherical equivalent prediction error (SEQ-PE)), precision (precision of SEQ-PE), accuracy (absolute SEQ-PE), and statistical analysis (robust two-sample t-test and Cochran's Q test)
Results
47 eyes were assessed with mean AL 23.88mm, ACD 2.34mm, LT 4.90mm, K 43.99D. EVO AL+K+ACD was the only formula with a trimmed mean within the top 3 formulae for each statistic (SEQ-PE:-0.032, Precision of SEQ-PE:0.368, Absolute SEQ-PE:0.366). For SEQ-PE, BUII Full and EVO AL+K+ACD was statistically different compared to others(P<0.05). for="" absolute="" seq-pe,="" evo="" al+k="" was="" statistically="" significant="" different="" to=""><0.05). subgroup="" analysis="" with="" fe="" measurements="" showed="" buii/evo="" formula="" improved="" with="" fe="" inputs="" but="" did="" not="" outperform="" al+k="" or="" al+k+acd="" formulae.="" for="" seq-pe,="" buii/evo="" al+k+acd="" was="" statistically="" significant="" different="" to="" buii=""><0.05). no="" other="" statistical="" differences="" were="">
Conclusion
IOL power calculations in zonulysis is challenging. Use of modern formulae performed well in our cohort. EVO AL+K+ACD performed best overall. Performance of BUII and EVO improved with ACD and LT inputs from the FE. Use of FE inputs however did not outperform BUII and EVO using AL+K or AL+K+ACD parameters.
Presenting Author
Woong-Joo Whang, MD
Co-Authors
Kyungmin Koh MD, PhD, MBA
Purpose
To investigate the accuracy of intraocular lens calculation formula using images obtained from swept-source optical coherence tomography(SS-OCT) and deep learning algorithm including convolutional neural network(CNN).
Methods
In this study, the preoperative image of the eye was measured using a SS-OCT. One type of intraocular lens was inserted in a total of 550 eyes. DenseNet, EfficientNet, InceptionV3, InceptionResNetV2, ResNet, VGG19, and Xception were used for IOL power calculation. Each CNN was used to train the images of 440 eyes, and then algorithm was tested in 110eyes. The prediction error of the Effective lens position (ELP) calculated according to the manifest refraction measured at 2 months postoperatively was measured.
Results
For the mean squared error of ELP, InceptionResNet and DenseNet were the most accurate, while for the R-squared value, InceptionResNet and ResNet were the most accurate.
Conclusion
Combining image information with various CNN algorithms may be a good alternative for IOL power calculation. In particular, InceptionResNet and DenseNet have shown accurate results.
Presenting Author
Dong Hyun Kim, MD, PhD
Purpose
Dry eye is known to be related with increased measurement variability in ocular biometry. This study aimed to investigate the short-term variability of ocular biometric measurements and to analyze the factors associated with high variability of biometric values in patients with cataract.
Methods
This retrospective study included 60 eyes of 60 patients who underwent ocular biometry twice with the IOLMaster 700 before cataract surgery, with an interval of 17.1�5.2 days. Ocular surface parameters included tear film break-up time (TBUT), corneal and conjunctival staining scores (CSS/ConSS). Refractive surprise was defined as a mean absolute error exceeding 0.5 diopter (D).
Results
Participants included 22 men and 38 women with a mean age of 70.6�8.4 years. The variability of mean keratometric (K) values and astigmatism showed notable variability compared to axial length. (mean K: 0.18�0.17D, astigmatism: 0.21�0.26D, axial length: 0.01�0.01mm) A decreased TBUT and increased CSS were significantly associated with high variability (?0.25D) of mean K. (p=0.040/0.008) The high K variability group (?0.25D) revealed a significantly higher incidence of refractive surprise (>0.5D) compared with the low K variability group (<0.25d). (45.5%/17.9%,="" p="">
Conclusion
Preoperative lower TBUT and higher CSS influenced the variability of K values in ocular biometry and may be related with postoperative refractive surprise due to erroneous K values.
Presenting Author
Xiteng Chen, MD
Purpose
Keratometry is crucial for intraocular lens power calculation in cataract surgery,with the IOL Master 700 offering reliable measurements in normal eyes. However, the extent to which tear film stability affects corneal refractive power measurement and surgical planning is not well-established, warranting further investigation in cataract patients.
Methods
Prospective cohort study. Using the results of Keratograph 5M, patients with cataracts preoperatively were divided into a dry-eye group (tear film stability level 2) and a control group (level 0-1). Two consecutive measurements were performed using IOL Master 700 with an interval of 10 minutes. The standard keratometry, keratometric corneal astigmatism (KCA), total keratometry and total corneal astigmatism (TCA) were recorded. The IOL power was calculated using SRK-T, SRK-T TK, Haigis, Haigis TK, Barrett Universal II, and Barrett Universal II TK formulas. The variability and the mean magnitude of corneal astigmatism vector difference were calculated and compared between groups.
Results
?Kf, ?KCA, ?TKf, ?TK, and vector variability of corneal astigmatism were significantly higher in the dry-eye group (all P < 0.05). Bland-Altman analyses showed broader 95% limits of agreement in the dry-eye group. Of all formulas, SRK-T displayed significantly higher variability in IOL power calculations in the dry eye group compared to the control group (P < 0.05).
Conclusion
Tear film instability can result in deviations in the measurement of preoperative corneal refractive power and corneal astigmatism. Dry eyes contribute to an increased prediction error in the IOL power calculation, especially for the SRK-T formula.
Presenting Author
Johnny L. Gayton, MD
Purpose
To compare the accuracy of astigmatic outcomes based on direct anterior and total corneal astigmatism measurements from a point source LED topographer versus keratometric values captured by swept source OCT in eyes undergoing cataract surgery.
Methods
Retrospective, single surgeon, single site, non-interventional study of 161 eyes that underwent previous toric IOL implantation following cataract surgery with preoperative biometry measurements using an LED topographer (Ambient; Cassini Technologies, B.V.) and swept source OCT biometer (Argos; Alcon Vision LLC.). The Ambient LED topographer captures a direct measurement of the total corneal astigmatism, whereas the Argos SS-OCT biometer provides a predictive posterior corneal astigmatic value. Residual cylinder, prediction errors, manifest refraction, and UCDVA, BCDVA were collected at one month postoperatively.
Results
In 99 eyes (Group A) data were obtained with both LED topography and SS-OCT; in 62 eyes (Group B), only SS-OCT data were collected and results were, respectively: mean residual cylinder 0.36�0.45D vs 0.43D�0.31D (p=0.22); mean prediction error -0.33�0.31D vs �0.34�0.45D (p=0.79); BCDVA at ?20/20, ?20/30, ?20/40 were 57%, 89%, and 100% for Group A vs 60%, 95%, and 98% for Group B (p=0.64). The percentage of eyes within 0.25D, 0.50D, and 0.75D were respectively measured as: 55.6%, 83.8%, and 95.7% in Group A vs 43.6%, 82.3%, and 91.9% in Group B. Within Group A, analysis of flat K values demonstrated that Ambient had lower SD vs Argos (1.78 vs 1.96) was statistically significant. (p=0.0029).
Conclusion
Results suggest that clinical astigmatic outcomes of Ambient LED topography are comparable to Argos SS-OCT biometry for toric IOL power calculation. However, the Ambient demonstrated significantly greater precision in measuring keratometric values compared to the Argos when measuring the same eyes.
Presenting Author
Sarah O Patterson, MD
Co-Authors
Davin Johnson MD, FRCSC, John Gonder MD
Purpose
Routine use of corneal and retinal/optic nerve imaging is increasingly standard of care for patients undergoing consultation for cataract surgery, with prior reports indicating high disease detection rates. Our study assessed the effectiveness of these screenings in a Canadian population of patients undergoing consultation for cataract surgery.
Methods
A retrospective chart review was conducted on 500 consecutive patient eyes evaluated for cataract surgery between 2019 and 2023 at an independent health facility in Ottawa, Ontario. Preoperative evaluation included Scheimpflug corneal tomography (Pentacam), OCT imaging of the optic nerve and retina, and clinical examination. Corneal tomography and macular OCT scans were reviewed by a cornea and retina specialist blinded to the examination results. The primary outcomes were the rate of clinically significant corneal and retinal abnormalities not previously detected or known before consultation. Secondary outcomes included the prevalence of clinically significant optic nerve head disease.
Results
Excluding patients with a history of corneal/refractive surgery, corneal imaging detected irregular astigmatism in 33.4% (140/419) of patient eyes, with the most common phenotype being superior steepening. For patients with OCT imaging, macular abnormalities were detected in 8.1% (37/458) of eyes, with 3/458 (0.7% of all eyes) having potential impact on surgical outcomes. In all three of these cases, the abnormality was detected only on OCT and not by clinical examination. In patients who had quality optic nerve head OCTs, 75.9% (289/381) of patient eyes had a normal average retinal nerve fiber layer thickness with 6.6% (25/381) of eyes having significant thinning.
Conclusion
Routine use of corneal and retinal imaging for cataract surgery patients is important for helping guide IOL choice and providing better prognostic information to patients. Nonetheless, our study found that the detection rate of clinically significant retinal disease, not otherwise detectable by clinical examination or prior diagnosis, was low.
Presenting Author
Andreas F. Borkenstein, MD
Purpose
Extreme axial eyeball length is a known risk factor for decreased accuracy of intraocular lens calculations. We report a case of a 35-year-old male with anisometropic amblyopia and an axial eyeball length of 35.7 mm.
Methods
After biometrics, various calculation formulas and methods were applied and compared. IOL power calculations were used, aiming for a slight postoperative residual myopia. Du to the extreme axial length, the patient required a minus-power intraocular lens. Moreover, due to the anatomical dimensions (large capsule) an IOL with a larger overall diameter was selected. The complex process prior to the surgery to determine the right lens model/type and power is shown here.
Results
Six months after surgery, the refractive error was -1.50 D -0.75 D x 179� with a corrected visual acuity of 20/63. The large IOL was stable in the capsular bag, and there were no postoperative complications. The patient was able to wear spectacles with correction in both eyes and reported significant improvement in binocular vision and quality of life.
Conclusion
Negative power IOLs typically have different optic configurations and require special considerations during IOL calculation. Care should be taken to avoid postoperative hyperopic refractive error. With this case report, we demonstrate that in rare cases of extreme, high myopia good results can be achieved with correct IOL power implantation.
Presenting Author
Ariana R Maleki, BSc
Co-Authors
Abhishek Soni BSc, Saxon Hancock MD, Ahmad Kheirkhah MD
Purpose
To calculate IOL power using Barrett formula, either predicted or measured posterior corneal astigmatism (PCA) can be used. Limited data exist on comparison of predicted versus measured PCA for multifocal IOLs. Herein, we compared the accuracy of Barrett formula for power calculation for a toric trifocal IOL using either predicted or measured PCA.
Methods
This retrospective study included 62 eyes which underwent uneventful cataract surgery and implantation of a toric trifocal IOL (PanOptix Toric, Alcon). Patients with prior keratorefractive surgery were excluded. Preoperatively, all patients had measurements with IOLMaster 700 and Pentacam. The Barrett Universal II formula was used to calculate the IOL power across 4 different settings: anterior corneal power from either IOLMaster or Pentacam, and PCA either as predicted by the formula or as measured by Pentacam. We then calculated the differences between the 1-month postoperative refraction and the anticipated refraction in all 4 settings for the IOL implanted in surgery.
Results
There were no significant differences between IOLMaster and Pentacam in front mean keratometry (P=0.52), front astigmatism magnitude (P=0.57), or front astigmatism axis (P=0.53). No statistically significant differences were found in the Mean Absolute Error (P=0.48) and the Mean Numerical Error (P=0.55) across the 4 settings. Likewise, no statistically significant differences were noted in percentage of eyes within 0.25 D (P=0.33), 0.50 D (P=0.29), and 0.75 D (P=0.09) of the target refraction across the 4 settings. The axial length did not have any significant effects on these outcome measures.
Conclusion
To calculate IOL power with Barrett formula for a toric trifocal IOL, no significant difference exists between using either predicted PCA by the formula or measured PCA by Pentacam. This can save time by using the default value provided by the formula (i.e., predicted PCA) and avoiding the need for additional calculations using measured PCA.
Presenting Author
Lisa M. Nijm, MD, JD, ABO
Purpose
To study the association between hyperosmolarity and ocular light scatter in a cataract surgery population.
Methods
In a prospective, observational study, contiguous 20 second ocular scatter index (OSI) scans were recorded in hyperosmolar (?320 mOsm/L) and normal subjects (<308 mosm/l)="" with="" cataract="" nuclear="" opacity="" .="" osi="" was="" measured="" at="" screening,="" baseline="" and="" 90="" days="" following="" surgery.="" along="" with="" symptoms="" of="" ocular="" surface="" disease,="" slit="" lamp="" examination="" included="" corneal="" staining="" (0�3),="" tear="" film="" breakup="" time="" (tbut)="" and="" evaluation="" of="" meibomian="" gland="" disease="" (mgd).="" an="" additional="" cohort="" of="" hyperosmolar="" subjects="" were="" measured="" for="" osi="" at="" screening,="" baseline,="" and="" 5,="" 10,="" 15,="" and="" 30="" minutes="" following="" instillation="" of="" 0.18%="" sodium="" hyaluronate="">
Results
Thirty-one eyes of 31 patients were included in the analysis. In the group with cataract nuclear opacity ?3 (n=21), there was a significant difference in postoperative OSI variation when comparing patients with or without hyperosmolarity post-surgery (0.65�0.30 [n=11] vs 0.33�0.11 [n=10], p=0.005), with variations in the order of 2�3 OSI in hyperosmolar eyes. These changes were not significant when patients were stratified using standard measurements, i.e., staining (p=0.891), TBUT (p=0.749), symptoms (p=0.719), or MGD status (p=0.852). In the hyperosmolar group (n=10), instillation of HA reduced OSI at 10, 15, and 30 minutes (p< />
Conclusion
Hyperosmolar eyes exhibited increased light scatter variation following cataract surgery compared with normal osmolarity eyes, which results in a 7-fold increase in post-surgical dissatisfaction. Therefore, hyperosmolarity may represent a key factor to address in order to improve patient outcomes and overall satisfaction after cataract surgery.
Presenting Author
Francisco J. Castro, MD
Co-Authors
Rub�n Hern�ndez MSc, Mart�n Puzo MD, Francisco Segura PhD, Sara Marquina Mart�n OD, Marta Jim�nez PhD
Purpose
To compare the refractive and astigmatic prediction accuracy among different intraocular lens (IOL) power calculation formulas in patients implanted with aberration-free monofocal toric IOL during cataract surgery.
Methods
This is a retrospective chart review of consecutive case records of patients who underwent cataract surgery with the implantation of an aberration-free aspheric monofocal toric IOL (enVista toric, MX60T) from February 2022 to November 2023 and had postoperative manifest refraction performed one month after the surgery. Outcome measures included the comparison of spherical equivalent (SEQ-PE) and astigmatic (RA-PE) predicted errors from various IOL power calculation formulas (Abulafia-Koch, Kane�s, EVO, Castrop�s, Hoffer QST, Hill-RBF, and Barrett toric IOL formula) for the implanted IOL power. (Statistical analysis was performed with Eyetemis Analysis Tool).
Results
131 eyes (56.5% women, 75.9�9.8 years) were assessed. The absolute RA-PE (accuracy) of the IOL formulas evaluated ranged from 0.35 D to 0.51 D. While the absolute RA-PE was comparable for the four formulas, Castrop�s IOL formula showed the worst prediction error with the statistically significant difference in the RA-PE when compared with Barrett, EVO, Hoffer�s QST, Kane and Abulafia Koch formulas (all p<0.05). for="" various="" iol="" formulas,="" the="" absolute="" spherical="" equivalent="" prediction="" error="" (seq-pe)="" was="" 0.16�0.22="" d.="" while="" the="" evo="" formula="" showed="" the="" least="" seq-pe="" compared="" to="" others,="" it="" was="" statistically="" significantly="" lower="" when="" compared="" with="" hoffer="" qst="" and="" abulafia-koch="" (both="" p="">
Conclusion
The astigmatic prediction accuracy of Abulafia-Koch, Kane, EVO, Hoffer QST, Hill-RBF, and Barrett toric IOL formulas was comparable. The spherical equivalent prediction accuracy of Barrett, EVO, and Kane was comparable, although the EVO formula showed the least prediction error among the various IOL formulas compared.
Presenting Author
Ruti Sella, MD
Co-Authors
Jacob Megreli MD, MPH, Olga Reitblat MHA, MD, Noa Kapelushnik MD, Eitan Livny MD, Guy Gindis BSc, Uri Elbaz MD
Purpose
Large Language Models (LLMs) are being increasingly utilized in medicine. We evaluated the effectiveness of LLM chatbots, including GPT-3.5, GPT-4 (OpenAI), Gemini (Google), and Copilot (Microsoft), as clinical decision-support tools in cataract surgery for astigmatic patients who are candidates for toric intraocular lens (IOL) implantation.
Methods
100 eyes of patients undergoing routine cataract surgery evaluations were retrospectively reviewed. Patients with astigmatism > 1.0D detected by the OA-2000 optical biometer (Tomey, Nagoya, Japan) underwent corneal tomography with the Pentacam HR (Oculus, Wetzlar, Germany). The outputs were anonymized and independently reviewed by two senior ophthalmologists to assess toric IOL suitability. Preoperative measurements were also analyzed by GPT-3.5, GPT-4, Gemini, and Copilot, and their recommendations for toric IOL suitability documented. Outcome measures included Interobserver agreement between the ophthalmologists and each chatbot and sensitivity, specificity, PPV, and NPV for each chatbot.
Results
A total of 116 eyes were analyzed in our study. ChatGPT-4.0 demonstrated an accuracy of 60% in determining suitability for Toric IOL, with an AUC of 0.62, a PPV of 80%, a NPV of 40%, a sensitivity of 58%, and a specificity of 66%. In comparison, GPT-3.5 achieved an accuracy of 56%, an AUC of 0.62, a PPV of 83%, an NPV of 39%, a sensitivity of 47%, and a specificity of 77% (P<0.05). Both Gemini and Copilot exhibited inferior performance and accuracy, with AUC values of 0.54 and 0.53, respectively (P>0.05), and all other test measures for these chatbots were statistically insignificant (P>0.05). Additionally, Copilot had difficulty accurately interpreting biometry and tomography values.
Conclusion
This pioneering study found ChatGPT4.0 and 3.5 to have fair and reasonable performance in assessing Toric IOL suitability, while others underperformed. With the rapid advancement of technology, LLMs are continuously improving in their capabilities, holding significant potential to enhance clinical care as decision-support tool for cataract surgeons
Presenting Author
Jascha A. Wendelstein, MD
Co-Authors
Achim Langenbucher PhD, David Cooke MD, Damien Gatinel PhD, MD, Kamran Riaz MD
Purpose
To investigate the influence of the corneal epithelium on corneal power, particularly in special cases such as post-refractive surgery and keratoconus.
Methods
Retrospective observational study. Measurement data were obtained from a high-resolution anterior segment analyzer (MS-39, CSO, Scandicci, Italy). Corneal curvature and power data, as well as surface height data, were organized in a cylindrical coordinate system. Calculations considered one, two, and three refractive surfaces, examining the role of epithelial thickness and stromal curvature.
Results
The effect of the epithelium on corneal power was minimal in normal corneas (below 0.1 diopters), but it was considerable in keratoconus and post-refractive surgery cases, with differences up to 0.9 diopters. The effect decreased for larger measurement zones.
Conclusion
Incorporating epithelial thickness and stromal curvature into corneal power calculations is a crucial next step in accurate corneal power and intraocular lens calculation eyes with previous refractive surgery or keratoconus. This study highlights the need for advanced diagnostic and calculation methods in complex cases.
To log in, click the teal "Login" button in the upper right-hand corner of this page. If you are logged in but still do not have access, please check your 2025 Annual Meeting registration.
