ObjectiveTo assess the occurrence of CNV in patients presenting with flat irregular pigment epithelial detachments (FIPED). MethodsForty-five patients (49 eyes) with FIPED on OCT were enrolled in this retrospective study. There were 25 males (28 eyes) and 20 females (21 eyes). The mean age was 61.022±9.292 years. FFA, ICGA, spectral domain OCT and OCT angiography (OCTA) were performed in all patients during the same period. The FIPED was defined as an irregular elevation of the RPE allowing distinct visualization of Bruch’s membrane on OCT B-scan. The abnormal vascular signals from the deep retinal layer to the choroid layer on OCTA was defined as CNV. The CNV was classified into a type 1 CNV and a type 2 CNV according to the OCT characteristics. The CNV was classified into a typical and occult CNV according to the characteristics of the FFA image. Of all 49 eyes, fundus angiography revealed 18 eyes (36.7%) with CNV, and 31 eyes (63.3%) with no characteristic signs of CNV. FFA examination found that CNV in 8 eyes (classic CNV in 1 eyes, occult CNV in 7 eyes), which confirmed by OCT were type 1 CNV; transmitted fluorescence in 41 eyes. ICGA examination showed that CNV-like hyperfluorescence spots in 18 eyes, suspicious hyperfluorescence spots in late stage in 20 eyes, and choroidal high permeability in 11 eyes, respectively; and 18 CNV eyes were confirmed to be type 1 CNV by OCT. To compare the detection of CNV by OCTA and fundus angiography. ResultsOf the 49 eyes with FIPED, OCTA detected 36 eyes (73.5%) of type 1 CNV, and full or partial strong reflex signals were seen in FIPED; 13 eyes (26.5%) were not associated with CNV, and some strong reflection signals were found in FIPED in 9 eyes, 4 eyes with weak reflection signal. The FFA was examined for 1, 7 eyes of the classic and occult CNV, which confirmed to be type 1 CNV by OCTA. Among the 18 eyes with CNV which detected by ICGA, OCTA also found type 1 CNV. Among the 20 eyes with ICGA’s late suspicious strong fluorescent spots, OCTA showed 17 eyes of type 1 CNV; in 11 eyes with high choroidal permeability, OCTA showed type 1 CNV in 1 eye. Among the 36 eyes with CNV which detected by OCT, there were SRD in 32 eyes, no SRD in 2 eyes and retinal interlamellar cavities in 2 eyes. ConclusionOCTA can detect 73.5% of FIPED eyes with CNV. Compared with traditional fundus angiography, OCTA has a higher detection rate of CNV under FIPED. The FIPED of the internal strong reflection signal has a certain diagnostic value for the type 1 CNV.
ObjectiveTo observe the efficacy and safety of urokinase arterial thrombolysis in the treatment of central retinal artery occlusion (CRAO) at different time window.MethodsA retrospective study. From January 2014 to November 2019, 157 eyes (157 CRAO patients) in the Xi’an People's Hospital (Xi’an Fourth Hospital) were included in the study. There were 120 males and 37 females, with the average age of 54.87±12.12 years. The mean onset time was 65.66±67.44 h. All patients were tested with BCVA using international standard visual acuity chart, and the results were converted into logMAR visual acuity record. The arm-retinal circulation time (A-Rct) and the filling time (FT) of retinal arterial trunk-terminal filling time were measured by FFA. The mean logMAR BCVA was 2.44±0.46, the mean A-Rct and FT were 27.72±9.78 and 13.58±14.92 s respectively. According to the time window, the patients were divided into the onset 3-72 h group and the onset 73-240 h group, which were 115 patients and 42 patients respectively. There were no statistically significant difference between the 3-72 h group and the 73-240 h group in age, A-Rct and LogMR BCVA before treatment (χ2=－0.197, －1.242, －8.990; P=0.844, 0.369, 0.369); the difference was statistically significant in FT comparison (χ2=－3.652, P=0.000). Urokinase artery thrombolytic therapy was performed at different time window of 3-24 h, 25-72 h, 73-96 h, 97-120 h, 121-240 h after the onset of onset. Age and A-Rct of patients with different treatment time windows were compared, and the differences were not statistically significant (χ2=6.588, 6.679; P=0.253, 0.246).In comparison of FT and logMAR BCVA, the difference was statistically significant (χ2 =30.150, 71.378; P=0.000, 0.000). FFA was rechecked 24 hours after treatment, BCVA was rechecked 30 days after treatment. The changes of A-Rct, FT and BCVA before and after treatment were compared and analyzed. The occurrence of adverse reactions during and after treatment were observed. The two groups of measurement data were compared. The t test was used for those with normal distribution and χ2 test was used for those with non-normal distribution. Spearman correlation analysis was used to analyze the correlation between onset time and the difference of A-Rct, FT shortening time and logMAR BCVA after treatment.ResultsAt 24 h after CRAO treatment, A-Rct and FT of 157 cases were 19.64±6.50 and 6.48±7.36 s respectively, which were significantly shorter than those before treatment, and the differences were statistically significant (χ2=－16.236, －14.703; P=0.000, 0.000). The logMAR BCVA at 30 d after treatment was 1.72±0.76, which was significantly higher than that before treatment. The difference was statistically significant (χ2=－14.460, P=0.000). After CRAO urokinase arterial thrombolysis at different time window, there were statistically significant differences in A-Rct shortening time, FT shortening time, and logMAR BCVA difference (χ2=12.408, 24.200, 104.388; P=0.030, 0.000, 0.000). There was no statistically significant difference between the 3-72 h group and the 73-240 h group (χ2 =－1.042, P=0.297) in shortening time of A-Rct after treatment. The difference of FT shortening time was statistically significant (χ2=－3.581, P=0.000). The difference of logMAR BCVA was statistically significant (χ2=－9.905, P=0.000). The results of Spearman correlation analysis showed that there was no correlation between the onset time and the shortening time of A-Rct and FT after treatment (rp=－0.040, －0.081; P=0.436, 0.115), and negative correlation with the logMAR BCVA difference (rp=－0.486, P=0.000). One case of intracranial hemorrhage occurred after treatment, and it improved after dehydration to reduce cerebral edema, scavenging free radicals and brain protection.ConclusionsUrokinase arterial thrombolytic therapy is effective for CRAO within time window of 3-240 h, A-Rct, FT and LogMRA BCVA are all improved. However, with the prolongation of thrombolytic therapy time window, the therapeutic effect of urokinase arterial thrombolytic therapy is decreased. The therapeutic effect of Urokinase arterial thrombolytic therapy was better within 72 h.
ObjectiveTo observe the diagnostic value of six classification intelligent auxiliary diagnosis lightweight model for common fundus diseases based on fundus color photography. MethodsA applied research. A dataset of 2 400 color fundus images from Nanjing Medical University Eye Hospital and Zhejiang Mathematical Medical Society Smart Eye Database was collected, which was desensitized and labeled by a fundus specialist. Of these, 400 each were for diabetic retinopathy, glaucoma, retinal vein occlusion, high myopia, age-related macular degeneration, and normal fundus. The parameters obtained from the classical classification models VGGNet16, ResNet50, DenseNet121 and lightweight classification models MobileNet3, ShuffleNet2, GhostNet trained on the ImageNet dataset were migrated to the six-classified common fundus disease intelligent aid diagnostic model using a migration learning approach during training as initialization parameters for training to obtain the latest model. 1 315 color fundus images of clinical patients were used as the test set. Evaluation metrics included sensitivity, specificity, accuracy, F1-Score and agreement of diagnostic tests (Kappa value); comparison of subject working characteristic curves as well as area under the curve values for different models. ResultCompared with the classical classification model, the storage size and number of parameters of the three lightweight classification models were significantly reduced, with ShuffleNetV2 having an average recognition time per sheet 438.08 ms faster than the classical classification model VGGNet16. All 3 lightweight classification models had Accuracy > 80.0%; Kappa values > 70.0% with significant agreement; sensitivity, specificity, and F1-Score for the diagnosis of normal fundus images were ≥ 98.0%; Macro-F1 was 78.2%, 79.4%, and 81.5%, respectively. ConclusionThe intelligent assisted diagnosis of common fundus diseases based on fundus color photography is a lightweight model with high recognition accuracy and speed; the storage size and number of parameters are significantly reduced compared with the classical classification model.