Identifying Medical Diagnoses and Treatable Diseases by Image-Based Deep Learning.

The implementation of clinical-decision support algorithms for medical imaging faces challenges with reliability and interpretability. Here, we establish a diagnostic tool based on a deep-learning framework for the screening of patients with common treatable blinding retinal diseases. Our framework utilizes transfer learning, which trains a neural network with a fraction of the data of conventional approaches. Applying this approach to a dataset of optical coherence tomography images, we demonstrate performance comparable to that of human experts in classifying age-related macular degeneration and diabetic macular edema. We also provide a more transparent and interpretable diagnosis by highlighting the regions recognized by the neural network. We further demonstrate the general applicability of our AI system for diagnosis of pediatric pneumonia using chest X-ray images. This tool may ultimately aid in expediting the diagnosis and referral of these treatable conditions, thereby facilitating earlier treatment, resulting in improved clinical outcomes. VIDEO ABSTRACT.

Validation of the Prognostic Usefulness of the Gene Expression Profiling Test in Patients with Uveal Melanoma.

PURPOSE: To validate the prognostic usefulness of gene expression profile (GEP) testing in patients with uveal melanoma. To determine whether combining tumor size with the GEP classification provides additional prognostic value. DESIGN: Retrospective analysis. PARTICIPANTS: Patients with a diagnosis of choroidal melanoma examined at Yale New Haven Hospital; University of California, San Diego; and Memorial Sloan Kettering Cancer Center. METHODS: Patients' demographic and clinical data and tumor characteristics were collected. Univariate and multivariate Cox hazard regression analysis were used to assess the association between tumor characteristics and GEP classification with metastasis as an outcome. MAIN OUTCOME MEASURES: Metastasis-free survival (MFS). RESULTS: Of the 337 individuals included in the study, 87 demonstrated metastases. The mean follow-up time was 37.2 (standard deviation [SD], 40.2) months for patients with metastases and 55.0 (SD, 49.3) months for those without metastases. Tumors of larger thickness and GEP class 2 (vs. class 1) were associated significantly with increased risk of metastasis. Tumor thickness showed better prognostic usefulness than GEP classification (Wald statistic, 40.7 and 24.2, respectively). Class 2 tumors with a thickness of 7.0 mm or more were associated with increased risk of metastasis than tumors with a thickness of < 7.0 mm (hazard ratio [HR], 3.23; 95% confidence interval [CI], 1.61-6.51), whereas class 1 tumors with a thickness of 9.0 mm or more were associated with increased risk of metastasis than tumors with a thickness of < 9.0 mm (HR, 2.07; 95% CI, 0.86-4.99). No difference in MFS was found between patients with class 1A tumors compared with those with class 1B tumors (P = 0.8). Patients with class 2 tumors showed an observed 5-year MFS of 47.5% (95% CI, 36.0%-62.8%). CONCLUSIONS: Tumor size was the most significant predictor of metastasis and provided additional prognostic value independent of GEP classification. In addition, rates of metastasis for class 2 tumors were lower than estimates reported by Castle Bioscience, and no difference in rates of metastasis were found between class 1A and 1B tumors. This indicates that tumor size should be accounted for when relying on GEP for prognostication and that patients with GEP class 1A or 1B tumors may benefit from the same metastatic surveillance protocols. FINANCIAL DISCLOSURE(S): The author(s) have no proprietary or commercial interest in any materials discussed in this article.

Deep Learning Estimation of 10-2 and 24-2 Visual Field Metrics Based on Thickness Maps from Macula OCT.

PURPOSE: To develop deep learning (DL) systems estimating visual function from macula-centered spectral-domain (SD) OCT images. DESIGN: Evaluation of a diagnostic technology. PARTICIPANTS: A total of 2408 10-2 visual field (VF) SD OCT pairs and 2999 24-2 VF SD OCT pairs collected from 645 healthy and glaucoma subjects (1222 eyes). METHODS: Deep learning models were trained on thickness maps from Spectralis macula SD OCT to estimate 10-2 and 24-2 VF mean deviation (MD) and pattern standard deviation (PSD). Individual and combined DL models were trained using thickness data from 6 layers (retinal nerve fiber layer [RNFL], ganglion cell layer [GCL], inner plexiform layer [IPL], ganglion cell-IPL [GCIPL], ganglion cell complex [GCC] and retina). Linear regression of mean layer thicknesses were used for comparison. MAIN OUTCOME MEASURES: Deep learning models were evaluated using R2 and mean absolute error (MAE) compared with 10-2 and 24-2 VF measurements. RESULTS: Combined DL models estimating 10-2 achieved R2 of 0.82 (95% confidence interval [CI], 0.68-0.89) for MD and 0.69 (95% CI, 0.55-0.81) for PSD and MAEs of 1.9 dB (95% CI, 1.6-2.4 dB) for MD and 1.5 dB (95% CI, 1.2-1.9 dB) for PSD. This was significantly better than mean thickness estimates for 10-2 MD (0.61 [95% CI, 0.47-0.71] and 3.0 dB [95% CI, 2.5-3.5 dB]) and 10-2 PSD (0.46 [95% CI, 0.31-0.60] and 2.3 dB [95% CI, 1.8-2.7 dB]). Combined DL models estimating 24-2 achieved R2 of 0.79 (95% CI, 0.72-0.84) for MD and 0.68 (95% CI, 0.53-0.79) for PSD and MAEs of 2.1 dB (95% CI, 1.8-2.5 dB) for MD and 1.5 dB (95% CI, 1.3-1.9 dB) for PSD. This was significantly better than mean thickness estimates for 24-2 MD (0.41 [95% CI, 0.26-0.57] and 3.4 dB [95% CI, 2.7-4.5 dB]) and 24-2 PSD (0.38 [95% CI, 0.20-0.57] and 2.4 dB [95% CI, 2.0-2.8 dB]). The GCIPL (R2 = 0.79) and GCC (R2 = 0.75) had the highest performance estimating 10-2 and 24-2 MD, respectively. CONCLUSIONS: Deep learning models improved estimates of functional loss from SD OCT imaging. Accurate estimates can help clinicians to individualize VF testing to patients.

Deep Learning Approaches Predict Glaucomatous Visual Field Damage from OCT Optic Nerve Head En Face Images and Retinal Nerve Fiber Layer Thickness Maps.

PURPOSE: To develop and evaluate a deep learning system for differentiating between eyes with and without glaucomatous visual field damage (GVFD) and predicting the severity of GFVD from spectral domain OCT (SD OCT) optic nerve head images. DESIGN: Evaluation of a diagnostic technology. PARTICIPANTS: A total of 9765 visual field (VF) SD OCT pairs collected from 1194 participants with and without GVFD (1909 eyes). METHODS: Deep learning models were trained to use SD OCT retinal nerve fiber layer (RNFL) thickness maps, RNFL en face images, and confocal scanning laser ophthalmoscopy (CSLO) images to identify eyes with GVFD and predict quantitative VF mean deviation (MD), pattern standard deviation (PSD), and mean VF sectoral pattern deviation (PD) from SD OCT data. MAIN OUTCOME MEASURES: Deep learning models were compared with mean RNFL thickness for identifying GVFD using area under the curve (AUC), sensitivity, and specificity. For predicting MD, PSD, and mean sectoral PD, models were evaluated using R2 and mean absolute error (MAE). RESULTS: In the independent test dataset, the deep learning models based on RNFL en face images achieved an AUC of 0.88 for identifying eyes with GVFD and 0.82 for detecting mild GVFD significantly (P < 0.001) better than using mean RNFL thickness measurements (AUC = 0.82 and 0.73, respectively). Deep learning models outperformed standard RNFL thickness measurements in predicting all quantitative VF metrics. In predicting MD, deep learning models based on RNFL en face images achieved an R2 of 0.70 and MAE of 2.5 decibels (dB) compared with 0.45 and 3.7 dB for RNFL thickness measurements. In predicting mean VF sectoral PD, deep learning models achieved high accuracy in the inferior nasal (R2 = 0.60) and superior nasal (R2 = 0.67) sectors, moderate accuracy in inferior (R2 = 0.26) and superior (R2 = 0.35) sectors, and lower accuracy in the central (R2 = 0.15) and temporal (R2 = 0.12) sectors. CONCLUSIONS: Deep learning models had high accuracy in identifying eyes with GFVD and predicting the severity of functional loss from SD OCT images. Accurately predicting the severity of GFVD from SD OCT imaging can help clinicians more effectively individualize the frequency of VF testing to the individual patient.

PREVALENCE OF MISMATCH REPAIR GENE MUTATIONS IN UVEAL MELANOMA.

PURPOSE: Uveal melanomas are associated with characteristic genetic changes. Germline mutations in mismatch repair (MMR) genes and microsatellite instability have been implicated in the development of numerous malignant neoplasms such as colon and ovarian cancers. The frequency of MMR defects in uveal melanomas has yet to be determined. METHODS: Here, we analyzed the frequency of MMR gene mutations in uveal melanoma specimens from the University of California, San Diego (UCSD), The Cancer Genome Atlas (TGCA), and the Catalogue of Somatic Mutations in Cancer (COSMIC). RESULTS: We identified only two mutations in a MMR gene: one premature stop codon in the PMS gene within the UCSD cohort (0.5% frequency) and one in-frame deletion in MSH3 within the COSMIC database (0.8% frequency). We report copy number variation of MLH1 in monosomy 3 and show decreased mRNA expression of MLH1 in uveal melanoma specimens with monosomy 3. Expression levels of MLH1 were not found to correlate with the observed number of total mutations. CONCLUSION: Overall, we show that mutations in MMR genes in uveal melanoma specimens are exceedingly rare, and although one copy of MLH1 is lost in monosomy 3, it does not seem to have pathologic consequences in uveal melanoma pathogenesis.

Comparison of conventional color fundus photography and multicolor imaging in choroidal or retinal lesions.

PURPOSE: Our purpose was to compare the characteristics of the retinal and choroidal lesions including choroidal nevus, choroidal melanoma and congenital hypertrophy of the retina pigment epithelium using conventional color fundus photography (CFP) and multicolor imaging (MCI). METHODS: The paired images of patients with retinal or choroidal lesions were assessed for the visibility of lesion's border, halo and drusen using a grading scale (0-2). The area of the lesion was measured on both imaging modalities. The same grading was also done on the individual color channels of MCI for a further evaluation. RESULTS: Thirty-three eyes of 33 patients were included. There were no significant differences in the mean border, drusen and halo visibility scores between the two imaging modalities (p = 0.12, p = 0.70, p = 0.35). However, the mean area of the lesion was significantly smaller on MCI than that on CFP (14.9±3.3 versus 18.7±3.4 mm2, p = 0.01). CONCLUSION: The appearance of choroidal and/ or retinal lesions on MCI may be different than that on CFP. Though MCI can provide similar information with CFP for the features of retinal and/ or choroidal lesions including border, halo and drusen; the infrared light reflection on MCI underestimates the extent of the choroidal lesion by 33%.

Ophthalmic manifestations of tuberous sclerosis: a review.

Tuberous sclerosis or tuberous sclerosis complex (TSC), one of the phakomatoses, is characterized by hamartomas of the heart, kidney, brain, skin and eyes. Ophthalmologic examinations are required in all cases of TSC. Retinal hamartomas are the most common ocular finding in tuberous sclerosis. The majority of hamartomas are non-progressive; however, lesions with subretinal fluid and progression have been reported. This paper details the genetics, clinical features and ocular findings of TSC and reviews potential therapeutic options for ophthalmic manifestations.

Detecting glaucomatous change in visual fields: Analysis with an optimization framework.

Detecting glaucomatous progression is an important aspect of glaucoma management. The assessment of longitudinal series of visual fields, measured using Standard Automated Perimetry (SAP), is considered the reference standard for this effort. We seek efficient techniques for determining progression from longitudinal visual fields by formulating the problem as an optimization framework, learned from a population of glaucoma data. The longitudinal data from each patient's eye were used in a convex optimization framework to find a vector that is representative of the progression direction of the sample population, as a whole. Post-hoc analysis of longitudinal visual fields across the derived vector led to optimal progression (change) detection. The proposed method was compared to recently described progression detection methods and to linear regression of instrument-defined global indices, and showed slightly higher sensitivities at the highest specificities than other methods (a clinically desirable result). The proposed approach is simpler, faster, and more efficient for detecting glaucomatous changes, compared to our previously proposed machine learning-based methods, although it provides somewhat less information. This approach has potential application in glaucoma clinics for patient monitoring and in research centers for classification of study participants.

Proactive Decision Support for Glaucoma Treatment: Predicting Surgical Interventions with Clinically Available Data.

A longitudinal ophthalmic dataset was used to investigate multi-modal machine learning (ML) models incorporating patient demographics and history, clinical measurements, optical coherence tomography (OCT), and visual field (VF) testing in predicting glaucoma surgical interventions. The cohort included 369 patients who underwent glaucoma surgery and 592 patients who did not undergo surgery. The data types used for prediction included patient demographics, history of systemic conditions, medication history, ophthalmic measurements, 24-2 VF results, and thickness measurements from OCT imaging. The ML models were trained to predict surgical interventions and evaluated on independent data collected at a separate study site. The models were evaluated based on their ability to predict surgeries at varying lengths of time prior to surgical intervention. The highest performing predictions achieved an AUC of 0.93, 0.92, and 0.93 in predicting surgical intervention at 1 year, 2 years, and 3 years, respectively. The models were also able to achieve high sensitivity (0.89, 0.77, 0.86 at 1, 2, and 3 years, respectively) and specificity (0.85, 0.90, and 0.91 at 1, 2, and 3 years, respectively) at an 0.80 level of precision. The multi-modal models trained on a combination of data types predicted surgical interventions with high accuracy up to three years prior to surgery and could provide an important tool to predict the need for glaucoma intervention.

Refractive Changes After Implantation of Reversed Intraocular Lens in Cataract Surgery: A Mathematical Model.

PURPOSE: To develop a mathematical model that can predict the amount of refractive change caused by implantation of an intraocular lens (IOL) in a reversed position during cataract surgery. METHODS: A theoretical mathematical formula based on the Gullstrand eye model was constructed to estimate the refractive change of the eye after implantation of a reversed IOL. The refractive change caused by implantation of the IOL in a reversed position was calculated based on the exchange of the anterior curvature with the posterior curvature of the IOL, and the lengthening of the distance between the IOL and the retina. In case of a three-piece IOL with angulation, the amount of refractive change was calculated based on its angle and the total refractive power of the eye, which is dependent on the focal length of the eye. RESULTS: Calculated refractive change for one-piece IOLs was less than 0.10 diopter (D). For three-piece IOLs, the calculated refractive change makes the eye on average 0.77 D more myopic and can increase with the total refractive power of the patient's eye. The mathematical model was applied to seven previously published cases of reverse IOL implantation. CONCLUSIONS: This calculation demonstrates that with an upside-down IOL, there is a small refractive change in the one-piece IOL, including a toric IOL without angulation, but there can be a large refractive change in the three-piece IOL with angulation, especially using a higher power IOL or with a shorter axial length. [J Refract Surg. 2023;39(5):326-331.].

Retinal vein occlusion is associated with stroke independent of underlying cardiovascular disease.

BACKGROUND: Stroke is a leading cause of mortality and morbidity. Thus, identifying associated risk factors may lead to earlier interventions aimed at reducing the risk of stroke development. Since cardiovascular disease simultaneously increases the risk of stroke and retinal vein occlusion (RVO), we sought to determine whether RVO is associated with the risk of stroke independent of underlying cardiovascular co-morbidities. METHODS: In this cross-sectional study, we reviewed the records of 80,754 individuals who were evaluated by an ophthalmologist over a 6-year period. We identified individuals with RVO, stroke and cardiovascular diseases including hypertension, diabetes mellitus, carotid disease, coronary artery disease and atrial fibrillation. Multivariable logistic regression models were used to analyze odds ratios for RVO and stroke. RESULTS: After adjusting for age, sex, cardiovascular disease and other risk factors, we found that the presence of RVO was associated with an odds ratio for stroke of 1.73 (CI, 1.40-2.12, p < 0.001). The association between RVO and stroke, after adjusting for sex and cardiovascular co-morbidities, was significantly stronger in individuals younger than 50 years of age, with an odds ratio of having a stroke of 3.06 (1.34-6.25, p < 0.001), while the presence of RVO in individuals older than 85 years was not significantly associated with stroke 1.19 (0.77-1.79, p = 0.41). CONCLUSIONS: Our findings demonstrate that RVO is significantly associated with stroke, even after adjusting for underlying cardiovascular co-morbidities. This association was highly significant in younger subjects, while not significant in older individuals.

32-year-old diabetic patient with progressive vision loss and crystalline retinopathy.

PURPOSE: To report the case of severe bilateral retinal vascular occlusion in a patient with hyperoxalosis and chronic renal failure. METHODS: Observational case report. Medical and imaging records were retrospectively reviewed. The patient was imaged with ultra-widefield (UWF) fundus photography and fluorescein angiography (UWF-FA), cross sectional and en face spectral-domain optical coherence tomography (SD-OCT), and OCT angiography. RESULTS: A 32-year-old diabetic patient receiving peritoneal dialysis was referred because of severe vision loss. UWF color fundus photography showed diffuse sclerotic retinal vessels and diffuse intraretinal crystals in both eyes. UWF-FA illustrated near-complete retinal vascular occlusion and capillary wipe out in both eyes. SD-OCT demonstrated diffuse inner and middle retina thinning in both eyes and multiple intraretinal hyperreflective foci consistent with crystalline deposits in all retina layers of both eyes. OCT angiography revealed severe capillary and large vessel non-perfusion in the superficial and deep retinal capillary plexus of each eye. The serum oxalate levels were increased at 28 µmol/L (reference range < 2 µmol/L) and genetic testing was positive for a heterozygous mutation of the AGXT (Alanine-Glyoxylate Amino Transferase) gene that causes type 1 autosomal recessive primary hyperoxaluria. CONCLUSION: A diagnosis of hyperoxalosis causing severe retinal vascular occlusion was rendered. Hyperoxalosis was the result of multiple factors including heterozygous AGXT mutation, chronic renal failure insufficiently treated with peritoneal dialysis, and a diet high in oxalate. This case highlights the importance of ruling out retinal oxalosis in patients on peritoneal dialysis in order to initiate prompt hemodialysis and prevent severe retinal vascular occlusion.

Detecting Glaucoma from Fundus Photographs Using Deep Learning without Convolutions: Transformer for Improved Generalization.

Purpose: To compare the diagnostic accuracy and explainability of a Vision Transformer deep learning technique, Data-efficient image Transformer (DeiT), and ResNet-50, trained on fundus photographs from the Ocular Hypertension Treatment Study (OHTS) to detect primary open-angle glaucoma (POAG) and identify the salient areas of the photographs most important for each model's decision-making process. Design: Evaluation of a diagnostic technology. Subjects Participants and Controls: Overall 66 715 photographs from 1636 OHTS participants and an additional 5 external datasets of 16 137 photographs of healthy and glaucoma eyes. Methods: Data-efficient image Transformer models were trained to detect 5 ground-truth OHTS POAG classifications: OHTS end point committee POAG determinations because of disc changes (model 1), visual field (VF) changes (model 2), or either disc or VF changes (model 3) and Reading Center determinations based on disc (model 4) and VFs (model 5). The best-performing DeiT models were compared with ResNet-50 models on OHTS and 5 external datasets. Main Outcome Measures: Diagnostic performance was compared using areas under the receiver operating characteristic curve (AUROC) and sensitivities at fixed specificities. The explainability of the DeiT and ResNet-50 models was compared by evaluating the attention maps derived directly from DeiT to 3 gradient-weighted class activation map strategies. Results: Compared with our best-performing ResNet-50 models, the DeiT models demonstrated similar performance on the OHTS test sets for all 5 ground-truth POAG labels; AUROC ranged from 0.82 (model 5) to 0.91 (model 1). Data-efficient image Transformer AUROC was consistently higher than ResNet-50 on the 5 external datasets. For example, AUROC for the main OHTS end point (model 3) was between 0.08 and 0.20 higher in the DeiT than ResNet-50 models. The saliency maps from the DeiT highlight localized areas of the neuroretinal rim, suggesting important rim features for classification. The same maps in the ResNet-50 models show a more diffuse, generalized distribution around the optic disc. Conclusions: Vision Transformers have the potential to improve generalizability and explainability in deep learning models, detecting eye disease and possibly other medical conditions that rely on imaging for clinical diagnosis and management.

Retinal Ischemic Perivascular Lesions in Individuals With Atrial Fibrillation.

Background We previously demonstrated that retinal ischemic perivascular lesions (RIPLs), which are indicative of ischemia in the middle retina, may be a biomarker of ischemic cardiovascular disease. In this study, we sought to determine the relationship between RIPLs and atrial fibrillation, a common source of cardiac emboli. Methods and Results In this case-control study, we identified individuals between the ages of 50 and 90 years who had undergone macular spectral domain optical coherence tomography imaging. Individuals with atrial fibrillation were identified, and age- and sex-matched individuals from the same pool, but without a diagnosis of atrial fibrillation, were selected as controls. Spectral domain optical coherence tomography scans were reviewed by 3 independent and masked observers for presence of RIPLs. The relationship between RIPLs and atrial fibrillation was analyzed using multivariable logistic regression models. There were 106 and 91 subjects with and without atrial fibrillation, respectively. The percentage of subjects with RIPLs was higher in the atrial fibrillation group compared with the control group (57.5% versus 37.4%; P=0.005). After adjusting for age, sex, smoking history, hypertension, diabetes, coronary artery disease, carotid stenosis, stroke, and myocardial infarction, the presence of RIPLs was significantly associated with atrial fibrillation, with an odds ratio of 1.91 (95% CI, 1.01-3.59). Conclusions RIPLs are significantly associated with atrial fibrillation, independent of underlying ischemic heart disease or cardiovascular risk factors. This association may inform the diagnostic cardiovascular workup for individuals with RIPLs incidentally detected on optical coherence tomography scan of the macula.

The Sensitivity of Ultra-Widefield Fundus Photography Versus Scleral Depressed Examination for Detection of Retinal Horseshoe Tears.

PURPOSE: Ultra-widefield (UWF) imaging is commonly used in ophthalmology in tandem with scleral depressed examinations (SDE) to evaluate peripheral retinal disease. Because of the increased reliance on this technology in tele-ophthalmology, it is critical to evaluate its efficacy for detecting the peripheral retina when performed in isolation. Therefore, we sought to evaluate UWF imaging sensitivity in detecting retinal horseshoe tears (HSTs). STUDY DESIGN: Retrospective clinical validity and reliability study. METHODS: A single-institutional retrospective analysis was performed on patients at the Shiley Eye Institute, University of California, San Diego. Patients with HSTs seen on SDE who underwent treatment with laser were included in the study. A total of 140 patients with HSTs in the right and/or left eyes met the inclusion criteria. Those with concomitant ruptured globes, retinal detachments, and vitreous hemorrhages were excluded. A total of 123 patients with 135 HSTs were included in the final analysis. The primary outcome was the number of HSTs detected by UWF imaging. A secondary outcome was HST location. Sensitivity was measured with respect to HST location, and statistical significance was calculated by Fisher exact testing. RESULTS: A total of 69 (51.1%) HSTs were visualized on UWF images and 66 (48.9%) were not visualized. The sensitivity of UWF imaging in capturing HSTs was 7 of 41 (17.1%), 8 of 25 (32.0%), 7 of 14 (50.0%), and 47 of 55 (85.5%) for the superior, inferior, nasal, and temporal quadrants, respectively. Sensitivities between HST visibility and location were statistically significant (P < .001). CONCLUSIONS: Nearly half of HSTs were missed by UWF imaging. This study demonstrates that UWF imaging alone is not sufficiently sensitive to exclude the presence of HSTs.

Clinical Outcomes Comparison of Combined Small Incision Lenticule Extraction with Collagen Cross-Linking Versus Small Incision Lenticule Extraction Only.

Purpose: To evaluate clinical outcome during 24 months follow-up between small incision lenticule extraction combined with cross-linking (SMILE Xtra) and small incision lenticule extraction (SMILE) only. Setting. Ophthalmology Division of San Rossore Medical Center, Pisa, Italy. Design: Retrospective comparative case series. Methods: The study comprised 70 eyes (35 patients); 40 eyes were corrected using SMILE and 30 eyes were corrected using SMILE Xtra using a low energy protocol. The outcomes were compared at 1, 6, 12, and 24 months postoperatively. Results: The mean spherical equivalent (SEQ) reduced from -7.18 ± 1.21 D to -0.01 ± 0.09 D in the SMILE group and from -6.20 ± 2.99 D to -0.04 ± 0.1 D postoperatively in SMILE Xtra (p < 0.05). At 24 months the mean SEQs were -0.01 ± 0.24 D for SMILE and -0.15 ± 0.33 D for SMILE Xtra (p > 0.05). At 1, 6, 12, and 24 months, there were no statistically significant differences between the SMILE and SMILE Xtra groups in logarithm of the minimum angle of resolution (logMAR) uncorrected distance visual acuity (UDVA), safety, and efficacy index (p > 0.05). The mean average keratometry (K-avg) at 1, 6, 12, and 24 months after surgery did not shown any statistically significant difference between SMILE and SMILE Xtra group (p > 0.05). The mean maximum keratometry (K-max) readings at 1, 6, 12, and 24 months were not statistically significant between SMILE and SMILE Xtra group (p > 0.05). The preoperative mean thinnest point pachymetry (TTP) was 543.90 ± 22.85 µm in the SMILE group and 523.40 ± 37.01 µm in the SMILE Xtra group (p < 0.05). At 1, 6, 12, and 24 months the mean TTP was not statistically significant between the SMILE and SMILE Xtra groups (p > 0.05). At 24 months, the TTP was 408.29 ± 38.75 µm for the SMILE group and 402.22 ± 37 µm for the SMILE Xtra group (p > 0.05). In the preoperative period, the mean maximum posterior elevation (MPE) was 8.63 ± 4.35 µm for SMILE and 8.13 ± 2.54 µm for SMILE Xtra (p > 0.05). After the surgical procedure, both groups showed a statistically significant increase of the MPE (p < 0.05). At 24 months, the MPE was 11.00 ± 4.72 µm for SMILE Xtra and 10.14 ± 3.85 µm for the SMILE group (p > 0.05). In the preoperative period, the means of the root mean square (RMS) of high-order aberration (HOA) were 0.08 ± 0.03 µm for the SMILE group and 0.08 ± 0.03 µm for the SMILE Xtra group (p > 0.05). At 24 months, the RMS of HOA was 0.13 ± 0.07 µm for the SMILE group and 0.14 ± 0.07 µm for the SMILE Xtra group (p > 0.05). In the preoperative period, the root mean square of coma aberration (RMS-Coma) aberration was 0.06 ± 0.09 µm for the SMILE group and 0.04 ± 0.03 µm for the SMILE Xtra group (p > 0.05). At 24 months, the coma aberration of SMILE group was 0.12 ± 0.21 µm and 0.16 ± 0.25 µm for SMILE Xtra group (p > 0.05). Conclusions: SMILE Xtra procedure is a safe and simple procedure that can be offered to patients with high corneal ectasia risk because there were no differences in the indices of ectasia compared to the group treated only with SMILE which has a low corneal ectatic risk.

Deep Learning Image Analysis of Optical Coherence Tomography Angiography Measured Vessel Density Improves Classification of Healthy and Glaucoma Eyes.

PURPOSE: To compare convolutional neural network (CNN) analysis of en face vessel density images to gradient boosting classifier (GBC) analysis of instrument-provided, feature-based optical coherence tomography angiography (OCTA) vessel density measurements and OCT retinal nerve fiber layer (RNFL) thickness measurements for classifying healthy and glaucomatous eyes. DESIGN: Comparison of diagnostic approaches. METHODS: A total of 130 eyes of 80 healthy individuals and 275 eyes of 185 glaucoma patients with optic nerve head (ONH) OCTA and OCT imaging were included. Classification performance of a VGG16 CNN trained and tested on entire en face 4.5 × 4.5-mm radial peripapillary capillary OCTA ONH images was compared to the performance of separate GBC models trained and tested on standard OCTA and OCT measurements. Five-fold cross-validation was used to test predictions for CNNs and GBCs. Areas under the precision recall curves (AUPRC) were calculated to control for training/test set size imbalance and were compared. RESULTS: Adjusted AUPRCs for GBC models were 0.89 (95% CI = 0.82, 0.92) for whole image vessel density GBC, 0.89 (0.83, 0.92) for whole image capillary density GBC, 0.91 (0.88, 0.93) for combined whole image vessel and whole image capillary density GBC, and 0.93 (0.91, 095) for RNFL thickness GBC. The adjusted AUPRC using CNN analysis of en face vessel density images was 0.97 (0.95, 0.99) resulting in significantly improved classification compared to GBC OCTA-based results and GBC OCT-based results (P ≤ 0.01 for all comparisons). CONCLUSION: Deep learning en face image analysis improves on feature-based GBC models for classifying healthy and glaucoma eyes.

Detecting Glaucoma in the Ocular Hypertension Study Using Deep Learning.

Importance: Automated deep learning (DL) analyses of fundus photographs potentially can reduce the cost and improve the efficiency of reading center assessment of end points in clinical trials. Objective: To investigate the diagnostic accuracy of DL algorithms trained on fundus photographs from the Ocular Hypertension Treatment Study (OHTS) to detect primary open-angle glaucoma (POAG). Design, Setting, and Participants: In this diagnostic study, 1636 OHTS participants from 22 sites with a mean (range) follow-up of 10.7 (0-14.3) years. A total of 66 715 photographs from 3272 eyes were used to train and test a ResNet-50 model to detect the OHTS Endpoint Committee POAG determination based on optic disc (287 eyes, 3502 photographs) and/or visual field (198 eyes, 2300 visual fields) changes. Three independent test sets were used to evaluate the generalizability of the model. Main Outcomes and Measures: Areas under the receiver operating characteristic curve (AUROC) and sensitivities at fixed specificities were calculated to compare model performance. Evaluation of false-positive rates was used to determine whether the DL model detected POAG before the OHTS Endpoint Committee POAG determination. Results: A total of 1147 participants were included in the training set (661 [57.6%] female; mean age, 57.2 years; 95% CI, 56.6-57.8), 167 in the validation set (97 [58.1%] female; mean age, 57.1 years; 95% CI, 55.6-58.7), and 322 in the test set (173 [53.7%] female; mean age, 57.2 years; 95% CI, 56.1-58.2). The DL model achieved an AUROC of 0.88 (95% CI, 0.82-0.92) for the OHTS Endpoint Committee determination of optic disc or VF changes. For the OHTS end points based on optic disc changes or visual field changes, AUROCs were 0.91 (95% CI, 0.88-0.94) and 0.86 (95% CI, 0.76-0.93), respectively. False-positive rates (at 90% specificity) were higher in photographs of eyes that later developed POAG by disc or visual field (27.5% [56 of 204]) compared with eyes that did not develop POAG (11.4% [50 of 440]) during follow-up. The diagnostic accuracy of the DL model developed on the optic disc end point applied to 3 independent data sets was lower, with AUROCs ranging from 0.74 (95% CI, 0.70-0.77) to 0.79 (95% CI, 0.78-0.81). Conclusions and Relevance: The model's high diagnostic accuracy using OHTS photographs suggests that DL has the potential to standardize and automate POAG determination for clinical trials and management. In addition, the higher false-positive rate in early photographs of eyes that later developed POAG suggests that DL models detected POAG in some eyes earlier than the OHTS Endpoint Committee, reflecting the OHTS design that emphasized a high specificity for POAG determination by requiring a clinically significant change from baseline.