Objective:To apply the multi-modal deep learning model to automatically classify the ultra-widefield fluorescein angiography (UWFA) images of diabetic retinopathy (DR).Methods:A retrospective study. From 2015 to 2020, 798 images of 297 DR patients with 399 eyes who were admitted to Eye Center of Renmin Hospital of Wuhan University and were examined by UWFA were used as the training set and test set of the model. Among them, 119, 171, and 109 eyes had no retinopathy, non-proliferative DR (NPDR), and proliferative DR (PDR), respectively. Localization and assessment of fluorescein leakage and non-perfusion regions in early and late orthotopic images of UWFA in DR-affected eyes by jointly optimizing CycleGAN and a convolutional neural network (CNN) classifier, an image-level supervised deep learning model. The abnormal images with lesions were converted into normal images with lesions removed using the improved CycleGAN, and the difference images containing the lesion areas were obtained; the difference images were classified by the CNN classifier to obtain the prediction results. A five-fold cross-test was used to evaluate the classification accuracy of the model. Quantitative analysis of the marker area displayed by the differential images was performed to observe the correlation between the ischemia index and leakage index and the severity of DR.Results:The generated fake normal image basically removed all the lesion areas while retaining the normal vascular structure; the difference images intuitively revealed the distribution of biomarkers; the heat icon showed the leakage area, and the location was basically the same as the lesion area in the original image. The results of the five-fold cross-check showed that the average classification accuracy of the model was 0.983. Further quantitative analysis of the marker area showed that the ischemia index and leakage index were significantly positively correlated with the severity of DR ( β=6.088, 10.850; P<0.001). Conclusion:The constructed multimodal joint optimization model can accurately classify NPDR and PDR and precisely locate potential biomarkers.

Objective:To explore the prognostic value of epidermal growth factor receptor (EGFR) co-mutation status in patients with advanced lung adenocarcinoma.Methods:Clinical data of patients with stage ⅢB-Ⅳ lung adenocarcinoma who were first diagnosed in the Department of Respiratory and Critical Care Medicine, First Affiliated Hospital of Hebei North University from January 2019 to December 2022 were collected prospectively. Patients were divided into EGFR mutation group ( n=82) and EGFR co-mutation group ( n=74) according to whether EGFR was combined with other gene mutations. The level of circulating tumor DNA (ctDNA) in peripheral blood was measured by real time fluorescence quantitative PCR. Objective response rate (ORR), disease control rate (DCR), the levels of ctDNA in peripheral blood, and progression-free survival (PFS) were compared between two groups of patients before and after 1 month of treatment. The univariate and multivariate analyses were conducted by Cox proportional hazards regression model. Results:In the EGFR mutation group, there were 45 cases of EGFR19 deletion mutation and 37 cases of EGFR21 mutation. In the EGFR co-mutation group, there were 41 cases of EGFR19 deletion mutation, 33 cases of EGFR21 mutation, 46 cases of TP53 mutation, 16 cases of RB1 mutation, 6 cases of PTEN mutation, 2 cases of MET amplification, 1 case of ERBB2 mutation, 1 case of KRAS mutation, 1 case of RET rearrangement, and 1 case of ALK rearrangement. There were statistically significant differences between the EGFR mutation group and the EGFR co-mutation group in the maximum tumor diameter ( χ2=5.04, P=0.025) and stage ( χ2=3.92, P=0.048). The ORRs of the two groups were 64.63% (53/82) and 37.84% (28/74), respectively, with a statistically significant difference ( χ2=11.19, P<0.001). The DCRs were 96.34% (79/82) and 86.49% (64/74), respectively, with a statistically significant difference ( χ2=4.95, P=0.026). The ctDNA levels in the EGFR mutation group and EGFR co-mutation group after one month of treatment decreased compared to before treatment[2.63 (1.83, 3.30) ng/μl vs. 4.73 (3.92, 5.49) ng/μl, Z=-7.06, P<0.001; 4.26 (2.26, 6.07) ng/μl vs. 5.28 (4.37, 6.09) ng/μl, Z=-5.15, P<0.001], the ctDNA levels in the EGFR co-mutation group were higher than those in the EGFR mutation group before treatment and after 1 month of treatment ( Z=-2.47, P=0.013; Z=-4.29, P<0.001). In the EGFR co-mutation group, the ctDNA levels in peripheral blood of patients who were effectively treated with targeted therapy decreased after 1 month of treatment compared to before treatment [(2.03±0.63) ng/μl vs. (3.92±0.82) ng/μl, t=42.94, P<0.001], the levels of ctDNA in peripheral blood of ineffectively treated patients before and after 1 month of treatment were higher than those of effectively treated patients [(5.84±0.57) ng/μl vs. (3.92±0.82) ng/μl, t=-11.91, P<0.001; (5.87±1.64) ng/μl vs. (2.03±0.63) ng/μl, t=-14.43, P<0.001]. The median PFS of the EGFR mutation group and the EGFR co-mutation group of patients were 10.4 and 8.3 months, respectively, with a statistically significant difference ( χ2=22.28, P<0.001). Univariate analysis suggested that the maximum tumor diameter ( HR=0.10, 95% CI: 0.06-0.16, P<0.001), performance status (PS) score ( HR=0.09, 95% CI: 0.06-0.15, P<0.001), stage ( HR=0.09, 95% CI: 0.05-0.14, P<0.001), pre-treatment ctDNA level ( HR=12.04, 95% CI: 8.21-17.65, P<0.001), ctDNA level after 1 month of treatment ( HR=3.75, 95% CI: 3.10-4.54, P<0.001) and EGFR co-mutations ( HR=2.21, 95% CI: 1.57-3.12, P<0.001) were found to be significant factors affecting the PFS of stage ⅢB-Ⅳ lung adenocarcinoma patients receiving targeted therapy; Multivariate analysis demonstrated that PS score ( HR=0.25, 95% CI: 0.14-0.47, P<0.001), stage ( HR=0.49, 95% CI: 0.24-0.98, P=0.044), pre-treatment ctDNA level ( HR=4.73, 95% CI: 3.08-7.28, P<0.001), ctDNA level after 1 month of treatment ( HR=2.15, 95% CI: 1.65-2.80, P<0.001), and EGFR gene co-mutation ( HR=2.26, 95% CI: 1.40-3.64, P<0.001) were independent risk factors for PFS in stage ⅢB-Ⅳ lung adenocarcinoma patients receiving targeted therapy. Conclusion:Both the EGFR mutation group and EGFR co-mutation group show a decrease in ctDNA levels after targeted therapy for one month compared to before treatment. The median PFS of EGFR co-mutation patients is shorter than that of patients with a single EGFR mutation. PS score, stage, ctDNA levels before and after treatment, and EGFR gene co-mutation are all independent factors affecting PFS in stage ⅢB-Ⅳ lung adenocarcinoma patients after targeted therapy.