Creating realistic anterior segment optical coherence tomography images using generative adversarial networks.

AIMS: To develop a generative adversarial network (GAN) capable of generating realistic high-resolution anterior segment optical coherence tomography (AS-OCT) images. METHODS: This study included 142 628 AS-OCT B-scans from the American University of Beirut Medical Center. The Style and WAvelet based GAN architecture was trained to generate realistic AS-OCT images and was evaluated through the Fréchet Inception Distance (FID) Score and a blinded assessment by three refractive surgeons who were asked to distinguish between real and generated images. To assess the suitability of the generated images for machine learning tasks, a convolutional neural network (CNN) was trained using a dataset of real and generated images over a classification task. The generated AS-OCT images were then upsampled using an enhanced super-resolution GAN (ESRGAN) to achieve high resolution. RESULTS: The generated images exhibited visual and quantitative similarity to real AS-OCT images. Quantitative similarity assessed using FID scored an average of 6.32. Surgeons scored 51.7% in identifying real versus generated images which was not significantly better than chance (p value >0.3). The CNN accuracy improved from 78% to 100% when synthetic images were added to the dataset. The ESRGAN upsampled images were objectively more realistic and accurate compared with traditional upsampling techniques by scoring a lower Learned Perceptual Image Patch Similarity of 0.0905 compared with 0.4244 of bicubic interpolation. CONCLUSIONS: This study successfully developed and leveraged GANs capable of generating high-definition synthetic AS-OCT images that are realistic and suitable for machine learning and image analysis tasks.

Enhancing the Automated Detection of Implantable Collamer Lens Vault Using Generative Adversarial Networks and Synthetic Data on Optical Coherence Tomography.

PURPOSE: To investigate the efficacy of incorporating Generative Adversarial Network (GAN) and synthetic images in enhancing the performance of a convolutional neural network (CNN) for automated estimation of Implantable Collamer Lens (ICL) vault using anterior segment optical coherence tomography (AS-OCT). METHODS: This study was a retrospective evaluation using synthetic data and real patient images in a deep learning framework. Synthetic ICL AS-OCT scans were generated using GANs and a secondary image editing algorithm, creating approximately 100,000 synthetic images. These were used alongside real patient scans to train a CNN for estimating ICL vault distance. The model's performance was evaluated using statistical metrics such as mean absolute percentage error (MAPE), mean absolute error (MAE), root mean squared error (RMSE), and coefficient of determination (R2) for the estimation of ICL vault distance. RESULTS: The study analyzed 4,557 AS-OCT B-scans from 138 eyes of 103 patients for training. An independent, retrospectively collected dataset of 2,454 AS-OCT images from 88 eyes of 56 patients, used prospectively for evaluation, served as the test set. When trained solely on real images, the CNN achieved a MAPE of 15.31%, MAE of 44.68 µm, and RMSE of 63.3 µm. However, with the inclusion of GAN-generated and algorithmically edited synthetic images, the performance significantly improved, achieving a MAPE of 8.09%, MAE of 24.83 µm, and RMSE of 32.26 µm. The R2 value was +0.98, indicating a strong positive correlation between actual and predicted ICL vault distances (P < .01). No statistically significant difference was observed between measured and predicted vault values (P = .58). CONCLUSIONS: The integration of GAN-generated and edited synthetic images substantially enhanced ICL vault estimation, demonstrating the efficacy of GANs and synthetic data in enhancing OCT image analysis accuracy. This model not only shows potential for assisting postoperative ICL evaluations, but also for improving OCT automation when data paucity is an issue. [J Refract Surg. 2024;40(4):e199-e207.].

Robotic-assisted radical prostatectomy is pushing the boundaries: a national survey of frailty using the national surgical quality improvement program.

Background: Robotic-assisted radical prostatectomy (RARP) has been found to be comparable and, in some cases, favorable to open surgical approaches, while being used in a frailer population. Objectives: We aimed to illustrate the trend in population frailty and compare morbidity and mortality postoperatively in patients who underwent RARP. Design and Methods: The National Surgical Quality Improvement Program data set was used to select patients who underwent RARP between the years 2011-2019. Age, frailty indicators, surgical characteristics, and perioperative morbidity and mortality were compared between the years 2011-2019 using the chi-square test (χ2) for categorical variables and the one-way analysis of variance (ANOVA) for continuous variables. Results: Our patient population consisted of 66,683 patients who underwent RARP. There was an increase in mean age and frailty indicated by an increase in 5-item frailty score ⩾2, metabolic syndrome index = 3, and American Society of Anesthesiologists' (ASA) class ⩾3 between the years 2011-2019 (p < 0.001). Whereas the rate of mortality and morbidity, indicated by postoperative Clavien-Dindo grade ⩾4 and major morbidity, remained the same over the same period (p > 0.264). Furthermore, operative time and length of stay decreased over the same period (p < 0.001). Conclusion: RARP is being performed on more frail patients, with no added morbidity or mortality.

The 5-item frailty index predicts 30-day morbidity and mortality in radical nephrectomy patients: A propensity matched analysis.

PURPOSE: To assess the ability of the 5-item frailty index (5-IFi) score to predict 30-day morbidity and mortality post-radical nephrectomy (RN). METHODS: ACS-NSQIP database was used to select patients who underwent RN from 2011 to 2020. 5-IFi score was calculated by assigning a point for each of the following comorbidities: chronic obstructive pulmonary disease or pneumonia, congestive heart failure, dependent functional status, hypertension, and diabetes. Patients were divided into 3 frailty groups 0, 1, and ≥2. Patient demographics, medical comorbidities, prolonged length of stay, and prolonged operative time were compared between different groups; mortality and morbidity using the Clavien-Dindo classification (CVD). Multivariable logistic regression models and propensity score matching were performed as a sensitivity analysis to control for possible confounders. RESULTS: Cohort consisted of 36,682 patients whereby 5-IFi class 0, 1, and ≥2 included 11,564 (31.5%), 16,571 (45.2%), and 8,547 (25.3%) patients respectively. A multivariable analysis and propensity score matching revealed that patients with 5-IFi classes 1 and ≥ 2 were more likely to have a prolonged length of stay (OR = 1.11, 1.3, respectively) and to mortality (OR = 1.85 for frailty class ≥2); in addition to CVD class 1 & 2 (OR = 1.51, OR = 1.13, respectively), and CVD ≥ 4 (OR = 1.41, 1.86, respectively) as compared to 5-IFi class 0 (P < 0.001). CONCLUSION: The 5-IFi score was found to be an independent predictor of prolonged length of stay, morbidity, and mortality after RN. This tool can play a major role in preoperative risk assessment and patient counseling based on individualized risks.

Surgical Complications Requiring Reoperation in Open Versus Minimally Invasive Radical Nephrectomy: A Contemporary Analysis of the National Surgical Quality Improvement Program.

BACKGROUND: To explore reoperation rates for different radical nephrectomy (RN) approaches that are experiencing a shift from open radical nephrectomy (ORN) toward minimally invasive surgery (MIS), we aimed to compare reoperation rates along with their culprit etiologies between the different types of surgical approaches for RN. METHODS: The national surgical quality improvement program dataset was used to select patients who underwent RN between the years 2012-2019. A 1:1 propensity score matched analysis was used. Reoperation rates and causes were then compared between open and MIS approach. RESULTS: The propensity matched cohort included 15,294 patients. Reoperation rates due to large bowel injury (0.01 vs. 0.14%), vascular injury (0.07 vs. 0.22%), and other abdominal (0.5 vs. 0.77%) were more common in ORN as compared to MIS (MIS vs. ORN, respectively, p < 0.034). Reoperation due to hernia (0.14 vs. 0.03%) was more common in MIS as compared to ORN (p = 0.027). No difference was seen for small bowel injury and incision/wound debridement. General reoperation (1.61 vs. 2.22%) and mortality (0.57 vs. 1.47%) were also more common in ORN as compared to MIS (p < 0.008). CONCLUSION: Reoperation due to large bowel injury, vascular injury, and other abdominal was more likely to occur in ORN. Whereas reoperation due to hernia was more likely to occur in MIS. Surgical approach was an independent risk factor for immediate and early reoperation rates in RN patients. These results could be used to counsel patients pre-operatively on possible surgical approaches and complications.