Differentiating Glaucomatous Optic Neuropathy From Non-glaucomatous Optic Neuropathies Using Deep Learning Algorithms.

PURPOSE: A deep learning framework to differentiate glaucomatous optic disc changes due to glaucomatous optic neuropathy (GON) from non-glaucomatous optic disc changes due to non-glaucomatous optic neuropathies (NGONs). DESIGN: Cross-sectional study. METHOD: A deep-learning system was trained, validated, and externally tested to classify optic discs as normal, GON, or NGON, using 2183 digital color fundus photographs. A Single-Center data set of 1822 images (660 images of NGON, 676 images of GON, and 486 images of normal optic discs) was used for training and validation, whereas 361 photographs from 4 different data sets were used for external testing. Our algorithm removed the redundant information from the images using an optic disc segmentation (OD-SEG) network, after which we performed transfer learning with various pre-trained networks. Finally, we calculated sensitivity, specificity, F1-score, and precision to show the performance of the discrimination network in the validation and independent external data set. RESULTS: For classification, the algorithm with the best performance for the Single-Center data set was DenseNet121, with a sensitivity of 95.36%, precision of 95.35%, specificity of 92.19%, and F1 score of 95.40%. For the external validation data, the sensitivity and specificity of our network for differentiating GON from NGON were 85.53% and 89.02%, respectively. The glaucoma specialist who diagnosed those cases in masked fashion had a sensitivity of 71.05% and a specificity of 82.21%. CONCLUSIONS: The proposed algorithm for the differentiation of GON from NGON yields results that have a higher sensitivity than those of a glaucoma specialist, and its application for unseen data thus is extremely promising.

Macular telangiectasia with bilateral obliterated capillaries: a case report.

BACKGROUND: We report an otherwise healthy adult with macular telangiectasia with aneurysms, ischemia, and obliterated capillaries in both eyes. METHODS: This is a case report with a brief literature review. CASE PRESENTATION: A 58-year-old Iranian woman presented with a gradual decrease in vision, with recent deterioration. Past medical history was unremarkable, and best-corrected visual acuity was 3/10 in both eyes. Multimodal imaging including fundus photo, fluorescein angiography, optical coherence tomography, and optical coherence tomography angiography was carried out. In macula of both eyes, parafoveal telangiectasia, occluded vessels, capillary dropouts, and aneurysms were observed. While there were a dense circinate exudation and edema in the macula of the right eye and a thin and disorganized inner retinal layer in the left eye, the outer retina was intact in both eyes. En face optical coherence tomography angiography revealed capillary blunting and rarefaction in both superficial and deep capillary plexuses of retina. CONCLUSION: Our case most probably represents a case of type 3 macular telangiectasia in the absence of any systemic association.

Keratin pearl cyst formation after traumatic implantation of an eyelash into the anterior chamber: A case report.

Purpose: We introduce a case with creamy white pearl-like keratin cysts in the anterior chamber after a penetrating injury associated with eyelash implantation. Observations: A 5-year-old girl presented with a history of penetrating corneal injury with a knife ten months ago. An eyelash was removed from the anterior chamber during her previous primary repair. Her parents complained about the presence of a white mass in her repaired eye, which had appeared about nine months after surgery. After the visco-expression of these solid masses, the histopathological evaluation revealed keratinous material surrounded by multinucleated giant cells. Conclusions and importance: In a case of post-traumatic implantation of eyelash into the anterior chamber, despite removing the cilia, cysts may develop, which suggests proliferating epithelial cells embedded within the anterior chamber.

Accuracy of different lens power calculation formulas in patients with phacomorphic glaucoma.

PURPOSE: The purpose of the study was to determine the most accurate formula for intraocular lens (IOL) power calculation among five currently used formulas in eyes with phacomorphic glaucoma (PG) undergoing cataract extraction surgery. MATERIALS AND METHODS: In this prospective interventional case series Patients diagnosed with PG were undergone uneventful phacoemulsification and IOL implantation. After 3 months, the refractive outcome for each formula was evaluated with mean prediction error (PE), mean absolute error (MAE), and the percentages of eyes within 0.25 D and 0.5 D of predicted error. RESULTS: Twenty-three patients completed the study. PEs were significantly different among the 5 formulas (P = 0.019), and Holladay I had the least error (-0.02 ± 1.11). Haigis formula had the highest hyperopic shift (0.37 ± 1.22), highest MAE (0.99 ± 0.78) and the lowest percentages of desired PEs, while the SRK II produced the greatest percentages. The overall differences in MAE between the 5 formulas were statistically insignificant (P = 0.547). CONCLUSION: In some extreme situations like patients with PG, lower generation of IOL power calculation formulas may still produce more acceptable refractive outcomes.

Measurement of retinal nerve fiber layer thickness with a deep learning algorithm in ischemic optic neuropathy and optic neuritis.

This work aims at determining the ability of a deep learning (DL) algorithm to measure retinal nerve fiber layer (RNFL) thickness from optical coherence tomography (OCT) scans in anterior ischemic optic neuropathy (NAION) and demyelinating optic neuritis (ON). The training/validation dataset included 750 RNFL OCT B-scans. Performance of our algorithm was evaluated on 194 OCT B-scans from 70 healthy eyes, 82 scans from 28 NAION eyes, and 84 scans of 29 ON eyes. Results were compared to manual segmentation as a ground-truth and to RNFL calculations from the built-in instrument software. The Dice coefficient for the test images was 0.87. The mean average RNFL thickness using our U-Net was not different from the manually segmented best estimate and OCT machine data in control and ON eyes. In NAION eyes, while the mean average RNFL thickness using our U-Net algorithm was not different from the manual segmented value, the OCT machine data were different from the manual segmented values. In NAION eyes, the MAE of the average RNFL thickness was 1.18 ± 0.69 µm and 6.65 ± 5.37 µm in the U-Net algorithm segmentation and the conventional OCT machine data, respectively (P = 0.0001).

Gastric Schwannoma: A Case Report and Review of the Literature for Gastric Submucosal Masses Distinction.

Schwannomas origin from Schwann cells sheath and generally are benign, slow-growing, and asymptomatic neoplasms which frequently appear in the head and neck. Although gastrointestinal schwannoma is really rare, the most affected organ in GI system is the stomach. Gastric schwannoma forms 0.2% of all gastric tumors. This neoplasm is always detected as a submucosal mass, the same as other gastrointestinal stromal tumors. Although these tumors have almost the same presentations, they are completely different at therapeutic options and prognoses. Hence, it is important to distinguish them apart and make an accurate diagnosis to optimize treatment outcomes. Herein, we report a case of 28-year-old woman with frequent vomiting and abdominal pain caused by 5 × 6 cm schwannoma in the antrum of the stomach. This is a rare case of gastric outlet obstruction due to a massive schwannoma. In addition, all other probable submucosal masses will be discussed at different aspects.

Undifferentiated Pleomorphic Sarcoma of Pancreas: A Case Report and Review of the Literature for the Last Updates.

The most prevalent type of soft tissue sarcoma is undifferentiated pleomorphic sarcoma (UPS) or previously known as malignant fibrous histiocytoma. It accounts over 20% of all soft tissue sarcomas and occurs most frequently in the extremities, trunk, and retroperitoneum. However, it has been rarely observed in the digestive system. Pancreas sarcoma represents less than 1% of all pancreatic tumors, and primary UPS of the pancreas is even rarer. It exhibits high recurrence and poor prognosis. In this case, a 72-year-old woman with a UPS tumor which was located in the pancreas head and neck without adhesion to the retroperitoneum will be discussed.

The role of S-methylisothiourea hemisulfate as inducible nitric oxide synthase inhibitor against kidney iron deposition in iron overload rats.

BACKGROUND: Iron dextran is in common use to maintain iron stores. However, it is potentially toxic and may lead to iron deposition (ID) and impair functions of organs. Iron overload can regulate the expression of inducible nitric oxide synthase (iNOS) in some cells that has an important role in tissue destruction. S-methylisothiourea hemisulfate (SMT) is a direct inhibitor of iNOS, and this study was designed to investigate the effect of SMT against kidney ID in iron overload rats. MATERIALS AND METHODS: 24 Wistar rats (male and female) were randomly assigned to two groups. Iron overloading was performed by iron dextran 100 mg/kg/day every other day for 2 weeks. In addition, during the study, groups 1 and 2 received vehicle and SMT (10 mg/kg, ip), respectively. Finally, blood samples were obtained, and the kidneys were prepared for histopathological procedures. RESULTS: SMT significantly reduced the serum levels of creatinine and blood urea nitrogen. However, SMT did not alter the serum levels of iron and nitrite, and the kidney tissue level of nitrite. Co-administration of SMT with iron dextran did not attenuate the ID in the kidney. CONCLUSION: SMT, as a specific iNOS inhibitor, could not protect the kidney from ID while it attenuated the serum levels of kidney function biomarkers.