Clean mucosal area detection of gastroenterologists versus artificial intelligence in small bowel capsule endoscopy.

Studies comparing the detection of clean mucosal areas in capsule endoscopy (CE) using human judgment versus artificial intelligence (AI) are rare. This study statistically analyzed gastroenterologist judgments and AI results. Three hundred CE video clips (100 patients) were prepared. Five gastroenterologists classified the video clips into 3 groups (≥75% [high], 50%-75% [middle], and < 50% [low]) according to their subjective judgment of cleanliness. Visualization scores were calculated using an AI algorithm based on the predicted visible area, and the 5 gastroenterologists' judgments and AI results were compared. The 5 gastroenterologists evaluated CE clip video quality as "high" in 10.7% to 36.7% and as "low" in 28.7% to 60.3% and 29.7% of cases, respectively. The AI evaluated CE clip video quality as "high" in 27.7% and as "low" in 29.7% of cases. Repeated-measures analysis of variance (ANOVA) revealed significant differences in the 6 evaluation indicators (5 gastroenterologists and 1 AI) (P < .001). Among the 300 judgments, 90 (30%) were consistent with 5 gastroenterologists' judgments, and 82 (91.1%) agreed with the AI judgments. The "high" and "low" judgments of the gastroenterologists and AI agreed in 95.0% and 94.9% of cases, respectively. Bonferroni's multiple comparison test showed no significant difference between 3 gastroenterologists and AI (P = .0961, P = 1.0000, and P = .0676, respectively) but a significant difference between the other 2 with AI (P < .0001). When evaluating CE images for cleanliness, the judgments of 5 gastroenterologists were relatively diverse. The AI produced a relatively universal judgment that was consistent with the gastroenterologists' judgements.

DEEP LEARNING-BASED PREDICTION OF OUTCOMES FOLLOWING NONCOMPLICATED EPIRETINAL MEMBRANE SURGERY.

PURPOSE: We used deep learning to predict the final central foveal thickness (CFT), changes in CFT, final best corrected visual acuity, and best corrected visual acuity changes following noncomplicated idiopathic epiretinal membrane surgery. METHODS: Data of patients who underwent noncomplicated epiretinal membrane surgery at Severance Hospital from January 1, 2010, to December 31, 2018, were reviewed. Patient age, sex, hypertension and diabetes statuses, and preoperative optical coherence tomography scans were noted. For image analysis and model development, a pre-trained VGG16 was adopted. The mean absolute error and coefficient of determination (R 2 ) were used to evaluate the model performances. The study involved 688 eyes of 657 patients. RESULTS: For final CFT, the mean absolute error was the lowest in the model that considered only clinical and demographic characteristics; the highest accuracy was achieved by the model that considered all clinical and surgical information. For CFT changes, models utilizing clinical and surgical information showed the best performance. However, our best model failed to predict the final best corrected visual acuity and best corrected visual acuity changes. CONCLUSION: A deep learning model predicted the final CFT and CFT changes in patients 1 year after epiretinal membrane surgery. Central foveal thickness prediction showed the best results when demographic factors, comorbid diseases, and surgical techniques were considered.

Semantic Segmentation Dataset for AI-Based Quantification of Clean Mucosa in Capsule Endoscopy.

Background and Objectives: Capsule endoscopy (CE) for bowel cleanliness evaluation primarily depends on subjective methods. To objectively evaluate bowel cleanliness, we focused on artificial intelligence (AI)-based assessments. We aimed to generate a large segmentation dataset from CE images and verify its quality using a convolutional neural network (CNN)-based algorithm. Materials and Methods: Images were extracted and divided into 10 stages according to the clean regions in a CE video. Each image was classified into three classes (clean, dark, and floats/bubbles) or two classes (clean and non-clean). Using this semantic segmentation dataset, a CNN training was performed with 169 videos, and a clean region (visualization scale (VS)) formula was developed. Then, measuring mean intersection over union (mIoU), Dice index, and clean mucosal predictions were performed. The VS performance was tested using 10 videos. Results: A total of 10,033 frames of the semantic segmentation dataset were constructed from 179 patients. The 3-class and 2-class semantic segmentation's testing performance was 0.7716 mIoU (range: 0.7031-0.8071), 0.8627 Dice index (range: 0.7846-0.8891), and 0.8927 mIoU (range: 0.8562-0.9330), 0.9457 Dice index (range: 0.9225-0.9654), respectively. In addition, the 3-class and 2-class clean mucosal prediction accuracy was 94.4% and 95.7%, respectively. The VS prediction performance for both 3-class and 2-class segmentation was almost identical to the ground truth. Conclusions: We established a semantic segmentation dataset spanning 10 stages uniformly from 179 patients. The prediction accuracy for clean mucosa was significantly high (above 94%). Our VS equation can approximately measure the region of clean mucosa. These results confirmed our dataset to be ideal for an accurate and quantitative assessment of AI-based bowel cleanliness.

Repeatability of Central Corneal Thickness Measurement Using Rotating Scheimpflug Camera in Dry and Normal Eyes.

PURPOSE: To compare the repeatability of central corneal thickness (CCT) measurement using the Pentacam between dry eyes and healthy eyes, as well as to investigate the effect of artificial tears on CCT measurement. METHODS: The corneal thicknesses of 34 patients with dry eye and 28 healthy subjects were measured using the Pentacam. One eye from each subject was assigned randomly to a repeatability test, wherein a single operator performed three successive CCT measurements time points-before and 5 min after instillation of one artificial teardrop. The repeatability of measurements was assessed using the coefficient of repeatability and the intraclass correlation coefficient. RESULTS: The coefficient of repeatability values of the CCT measurements in dry and healthy eyes were 24.36 and 10.69 µm before instillation, and 16.85 and 9.72 µm after instillation, respectively. The intraclass correlation coefficient was higher in healthy eyes than that of in dry eyes (0.987 vs. 0.891), and it had improved significantly in dry eyes (0.948) after instillation of one artificial teardrop. The CCT measurement fluctuated in dry eyes (repeated-measures analysis of variance, P<0.001), whereas no significant changes were detected in healthy eyes, either before or after artificial tear instillation. CONCLUSIONS: Central corneal thickness measurement is less repeatable in dry eyes than in healthy eyes. Artificial tears improve the repeatability of CCT measurements obtained using the Pentacam in dry eyes.

Comparison of Corneal Epithelial Remodeling After Femtosecond Laser-Assisted LASIK and Small Incision Lenticule Extraction (SMILE).

PURPOSE: To comparatively investigate changes in epithelial thickness between myopic femtosecond laser-assisted LASIK (FS-LASIK) and small incision lenticule extraction (SMILE). METHODS: This study compared the topographic epithelial thickness changes in 175 myopic eyes undergoing FS-LASIK (62 eyes) or SMILE (113 eyes). Epithelial thickness was obtained using spectral-domain optical coherence tomography before surgery and 1 and 3 months after surgery. Topographic epithelial thickness obtained by automatic algorithm and thickness variability (standard deviation over 17 imaged areas) was compared between two groups. Postoperative epithelial thickness changes were correlated with treatment parameters. RESULTS: For FS-LASIK, the mean epithelial thickness of the center zone (2 mm in diameter), paracenter (2 to 5 mm), and mid-periphery (5 to 6 mm) increased by 3.4, 4.3, and 2.1 µm, respectively, at 1 month and by 4.4, 5.1, and 2.9 µm, respectively, at 3 months. There was an increase of 2.5, 3.9, and 4.5 µm, respectively, at 1 month and 3.0, 4.2, and 4.9 µm, respectively, at 3 months following SMILE. The epithelial thickness did not change between 1 and 3 months postoperatively following SMILE, whereas it increased further after FS-LASIK. A larger increase of epithelial thickness was observed in the central zone at 3 months following FS-LASIK than SMILE, whereas the opposite was observed in the mid-periphery. The topographic thickness variability was greater after FS-LASIK than SMILE. Corneal epithelial thickening was proportional to the amount of myopia correction after both procedures. CONCLUSIONS: Topographic epithelial remodeling patterns differ following FS-LASIK or SMILE. Epithelial remodeling appears to stabilize more rapidly following SMILE than FS-LASIK. [J Refract Surg. 2017;33(4):250-256.].

Correlation of Sex and Myopia With Corneal Epithelial and Stromal Thicknesses.

PURPOSE: To measure the corneal epithelial thickness (CET) and stromal thickness (ST) in normal subjects, and to also evaluate their correlation with gender, age, and myopia severity. METHODS: A total of 162 normal subjects were enrolled; they were divided into 3 groups based on their refractive status: "low myopia" (-0.25 to -3.0 D), "moderate myopia" (-3.0 to -6.0 D), and "high myopia" (more than -6.0 D). The CET and total corneal thickness were obtained using optical coherence tomography, and the regional thicknesses and topographic variability were compared among the 3 myopia groups and between men and women. Furthermore, the correlation between the thickness measurements and gender, age, and refractive status was analyzed. RESULTS: The CET and ST of the central segment (2 mm in diameter) did not significantly differ between the 3 groups. In the paracentral and midperipheral zones (annuli 2-5 mm and 5-6 mm from the center), the CET in some areas, and the ST in most areas, differed significantly among the 3 groups. The CET values of men and women in the central segment were 54.6 ± 3.5 µm and 52.8 ± 3.5 µm, respectively (P = 0.003). Similar differences were observed in the paracentral and midperipheral zones. Multiple regression analysis suggested that the CET is related to gender, whereas the ST is associated with myopia severity. Age was not correlated with either the CET or ST. CONCLUSIONS: The CET is significantly greater in men than in women, and is not correlated with myopia severity, whereas the ST is inversely correlated with myopia severity, without significant gender difference.

Spherical aberration reduction in nuclear cataracts.

PURPOSE: To measure higher-order aberrations (HOAs) in nuclear cataracts and to investigate spherical aberration changes with nuclear cataract progression. METHODS: A total of 102 eyes of older subjects (>50 years) were classified based on the nuclear opalescence (NO) grading of the Lens Opacities Classification System III: Group 1 (< NO2, 35), 2 (NO2, 23), 3 (NO3, 23), and 4 (≥ NO 4:21 eyes). Wave front measurements were performed with a Hartmann-Shack aberrometer. To investigate lenticular spherical aberration, HOAs were examined in 28 eyes before and after phacoemulsification, followed by insertion of an aberration-free intraocular lens. The relationship between lens opacity grade and ocular and lenticular spherical aberrations were investigated. RESULTS: Mean spherical aberrations in groups 1, 2, 3, and 4 were 0.25 ± 0.10, 0.16 ± 0.13, 0.12 ± 0.15, and 0.10 ± 0.20 µm, respectively, and showed a significant difference with nuclear opacity grading (p = 0.001). The spherical aberration showed negative associations with nuclear opacity grading (r = -0.408, p < 0.001). The predominant change in HOAs after phacoemulsification was an increase in spherical aberration, and 86 % of lenticular spherical aberrations were presumed to have negative values. The lenticular spherical aberration was negatively correlated with nuclear opacity severity (r = -0.409, p = 0.031). CONCLUSIONS: Ocular spherical aberration decreases with the progression of nuclear cataracts due to the negative correlation between lenticular spherical aberration and nuclear opacity severity.

Atypical hereditary sensory and autonomic neuropathy type IV with neither mental retardation nor pain insensitivity.

Hereditary sensory and autonomic neuropathy type IV is an autosomal recessive disorder characterized by severe mental retardation and self-mutilation-related complications. Recently, we investigated a 16-year-old Korean boy with normal intelligence. He had preserved pain sensation but was suspected of having hereditary sensory and autonomic neuropathy type IV because of the recurrent bone fractures and painless joint destruction in the absence of any predisposing medical conditions. Genetic analysis of the NTRK1 gene revealed compound heterozygous mutations including c.851-33T>A and c.2303C>T (p.Pro768Leu) in the NTRK1 gene. The p.Pro768Leu mutation has been identified in 2 Japanese patients with a mild phenotype. Therefore, although it is rare, hereditary sensory and autonomic neuropathy type IV should be considered in patients with recurrent bone fractures and painless joint destruction who do not have any predisposing conditions even when they do not have typical clinical features such as mental retardation or pain insensitivity.

Macular infarction associated with reactive arthritis.

A 53-year-old woman visited the Department of Rheumatology with a chief complaint of a 3-day history of fever and chills and also presented with pain occurring in both knees at the time of outpatient visit. Based on rheumatologic and hematological lab studies, ultrasonography, and a needle aspiration biopsy of the articular cavity, the patient was diagnosed with reactive arthritis. On hospitalization day 3, consultation with the Department of Ophthalmology was requested regarding decreased visual acuity lasting for 3 days. Upon ophthalmologic examination, the corrected visual acuity was 0.1 in the right eye and 0.05 in the left eye. Upon slit lamp microscopy, there were no abnormal findings in the anterior segment. Upon fundus examination, however, there were yellow-white lesions in the macular area of both eyes. Fluorescein angiography was performed to assess the macular lesions, and the findings were suggestive of macular infarction in both eyes. Due to a lack of other underlying disease, a past surgical history, and a past history of drug administration, the patient was diagnosed with macular infarction in both eyes associated with reactive arthritis. To date, there have been no other such cases reported. In a patient with reactive arthritis, we experienced a case of macular infarction in both eyes, which occurred without association with a past history of specific drug use or underlying disease. Herein, we report our case, with a review of the literature.

Valsalva retinopathy associated with an oratorical contest.

A 17-year-old man presented to us with a chief complaint of decreased visual acuity accompanied by central scotoma. There was nothing unusual in his medical history other than a recent oratorical contest. At the time of initial diagnosis, the corrected visual acuity was 20/20 in the right eye and 20/100 in the left eye. No significant findings were apparent on ophthalmic evaluation. On fundoscopy, there was a dumbbell-shaped macular bleed with a well-defined margin in the left eye. The clinical course was closely monitored along with drug therapy. Four weeks post presentation, the pre-retinal hemorrhage had nearly resolved. On fluorescein angiography, no significant findings were observed. In the left eye, the corrected visual acuity had improved to 20/25. Valsalva retinopathy is a pathology that occurs when a sudden increase in intra-thoracic pressure or abdominal pressure occurs in an otherwise healthy person. Here we report a case of Valsalva retinopathy occurring following an oratorical contest along with a review of the relevant literature.

Brimonidine 0.2% versus brimonidine Purite 0.15%: prophylactic effect on iop elevation after Nd:YAG laser posterior capsulotomy.

AIMS: The aim of this study was to compare the prophylactic effect of brimonidine 0.2% versus brimonidine Purite 0.15% on intraocular pressure (IOP) increase after Nd:YAG laser posterior capsulotomy. METHODS: In this prospective, double-masked, randomized, controlled study, 106 patients (106 eyes) who underwent Nd:YAG laser posterior capsulotomy were allocated to a brimonidine 0.2% group (35 eyes), a brimonidine Purite 0.15% group (36 eyes), or a vehicle group (35 eyes). One (1) drop of brimonidine 0.2%, brimonidine Purite 0.15%, or vehicle was instilled 1 h preoperatively and 1 drop immediately after Nd:YAG laser posterior capsulotomy. IOPs were measured preoperatively and at 1, 2, 3, and 24 h postoperatively. RESULTS: Decreases in IOP from baseline ranged from 2.3 to 2.7 mmHg in the brimonidine 0.2% group and 2.2-2.5 mmHg in the brimonidine Purite 0.15% group (P < 0.05), whereas the vehicle group exhibited a rise in IOP. IOP elevations of less than 5 mmHg occurred in 22.9% of patients in the brimonidine 0.2% group, 27.8% in the brimonidine Purite 0.15% group, and 48.6% in the vehicle group. Spikes of IOP greater than 10 mmHg occurred in 2.9% of patients in the brimonidine 0.2% group, 2.8% in the brimonidine Purite 0.15% group, and 8.6% in the vehicle group. The incidence of IOP elevation was not statistically significant between the brimonidine 0.2% and the brimonidine Purite 0.15% groups (P < 0.05). CONCLUSIONS: Brimonidine 0.2% and brimonidine Purite 0.15% have similar efficacy in the prevention of IOP elevation after Nd:YAG laser posterior capsulotomy.

Development of parafoveal exudates and serous retinal detachment in a pregnant woman with tuberous sclerosis.

The authors describe an unusual case of a serous detachment of the fovea and decreased vision with parafoveal exudates with subsequent spontaneous resolution and return of central vision in a pregnant patient with tuberous sclerosis. To our knowledge, this is the first report of an ocular change during gestation in a patient with tuberous sclerosis.