Development and validation of LC-MS/MS method for amlexanox in rat plasma and its application in preclinical pharmacokinetics.

Amlexanox, an anti-inflammatory and anti-allergic agent, has been widely used clinically for the treatment of canker sores, asthma, and allergic rhinitis. Recently, amlexanox has received considerable attention in curing nonalcoholic fatty liver diseases and hepatitis virus infection. Herein, we first established a sensitive high-performance liquid chromatography-tandem mass spectrum (LC-MS/MS) method for the determination of amlexanox in rat plasma. Propranolol was used as the internal standard (IS). Using a simple protein precipitation method, the amlexanox and IS were separated with Capcell Pak C18 column (2.0 × 50 mm, 5 µm) and eluted with water and acetonitrile each containing 0.1% formic acid using gradient elution condition at a flow rate of 0.4 mL·min-1 . Amlexanox and IS were detected by a triple quadrupole mass in multiple reactive monitoring (MRM) under the transitions of m/z 299.2 → 281.2 and m/z 259.9 → 116.1 with positive electrospray ionization, respectively. The calibration curves of amlexanox were established with the range of 50 to 2000 ng·mL-1 (r2 > 0.99). The validation method consisted of selectivity, accuracy, precision, carryover effect, matrix effect, recovery, dilution effect, and stability. The fully validated method was successfully applied to the pharmacokinetic study of amlexanox in Wistar rats.

Comparative Pharmacokinetics and Tissue Distribution of M10 and Its Metabolite Myricetin in Normal and Dextran-Sodium-Sulfate-Induced Colitis Mice.

M10, a novel myricetin derivative, is an anti-inflammatory agent designed for treatment of colitis. Here, we aim to investigate its pharmacokinetic behavior and tissue distribution in a mouse model with colitis. Pharmacokinetics and tissue distribution of M10 and its metabolite myricetin were compared in normal mice and in dextran-sodium-sulfate (DSS)-induced colitis mice. The role of fecal microbiota was also analyzed during metabolism of M10 in vitro. After oral administration, M10 was very low in the plasma of both normal and diseased mice. However, both M10 and myricetin were mainly distributed in the gastrointestinal tract, including the stomach, colon and small intestine, in physiological and pathological conditions. Significantly, M10 and myricetin were found in higher levels in gastrointestinal tracts with inflamed tissues than in normal tissues of mice. An in vitro assay revealed that 80% of M10 was metabolized to myricetin via fecal microbiota. After oral administration, M10 was not absorbed into circulation but mainly distributed in the inflamed submucosal tissues of colitic mice, where it was metabolized into myricetin to prevent colitis development.

Deep learning-based detection and stage grading for optimising diagnosis of diabetic retinopathy.

AIMS: To establish an automated method for identifying referable diabetic retinopathy (DR), defined as moderate nonproliferative DR and above, using deep learning-based lesion detection and stage grading. MATERIALS AND METHODS: A set of 12,252 eligible fundus images of diabetic patients were manually annotated by 45 licenced ophthalmologists and were randomly split into training, validation, and internal test sets (ratio of 7:1:2). Another set of 565 eligible consecutive clinical fundus images was established as an external test set. For automated referable DR identification, four deep learning models were programmed based on whether two factors were included: DR-related lesions and DR stages. Sensitivity, specificity and the area under the receiver operating characteristic curve (AUC) were reported for referable DR identification, while precision and recall were reported for lesion detection. RESULTS: Adding lesion information to the five-stage grading model improved the AUC (0.943 vs. 0.938), sensitivity (90.6% vs. 90.5%) and specificity (80.7% vs. 78.5%) of the model for identifying referable DR in the internal test set. Adding stage information to the lesion-based model increased the AUC (0.943 vs. 0.936) and sensitivity (90.6% vs. 76.7%) of the model for identifying referable DR in the internal test set. Similar trends were also seen in the external test set. DR lesion types with high precision results were preretinal haemorrhage, hard exudate, vitreous haemorrhage, neovascularisation, cotton wool spots and fibrous proliferation. CONCLUSIONS: The herein described automated model employed DR lesions and stage information to identify referable DR and displayed better diagnostic value than models built without this information.

Combination of a novel microtubule inhibitor MBRI-001 and gemcitabine synergistically induces cell apoptosis by increasing DNA damage in pancreatic cancer cell lines.

Pancreatic cancer (PC) is a highly malignant cancer with poor prognosis. Although gemcitabine (GEM; 2',2'-difluoro-deoxycytidine) has been used as the first-line chemotherapeutic agent in PC treatment for decades, its limited efficacy remains a significant clinical issue, which may be resolved by GEM combination therapy. In this study, we aimed to investigate the anti-tumor effects of MBRI-001 in combination with GEM in BxPC-3 and MIA PaCa-2 human PC cell lines. In vitro and in vivo results indicate that MBRI-001 showed synergistic activity with GEM. GEM induced apoptosis by increasing DNA damage (phosphorylated core histone protein H2AX (γ-H2AX)), MBRI-001 activated mitochondrial-apoptotic pathway (cleaved poly-ADP ribose polymerase (PARP)). Thus, the combination of the two intensified both apoptosis and DNA damage and showed significantly superior anti-tumor activity compared to each agent alone. The adoption of combination of MBRI-001 with GEM may be beneficial as they act synergistically and thus, can be a potential therapeutic choice for improving the prognosis of PC patients in the future.

Structure-based design and synthesis of novel furan-diketopiperazine-type derivatives as potent microtubule inhibitors for treating cancer.

Plinabulin, a synthetic analog of the marine natural product "diketopiperazine phenylahistin," displayed depolymerization effects on microtubules and targeted the colchicine site, which has been moved into phase III clinical trials for the treatment of non-small cell lung cancer (NSCLC) and the prevention of chemotherapy-induced neutropenia (CIN). To develop more potent anti-microtubule and cytotoxic derivatives, the co-crystal complexes of plinabulin derivatives were summarized and analyzed. We performed further modifications of the tert-butyl moiety or C-ring of imidazole-type derivatives to build a library of molecules through the introduction of different groups for novel skeletons. Our structure-activity relationship study indicated that compounds 17o (IC50 = 14.0 nM, NCI-H460) and 17p (IC50 = 2.9 nM, NCI-H460) with furan groups exhibited potent cytotoxic activities at the nanomolar level against various human cancer cell lines. In particular, the 5-methyl or methoxymethyl substituent of furan group could replace the alkyl group of imidazole at the 5-position to maintain cytotoxic activity, contradicting previous reports that the tert-butyl moiety at the 5-position of imidazole was essential for the activity of such compounds. Immunofluorescence assay indicated that compounds 17o and 17p could efficiently inhibit microtubule polymerization. Overall, the novel furan-diketopiperazine-type derivatives could be considered as a potential scaffold for the development of anti-cancer drugs.

Synthesis of deuterium-enriched sorafenib derivatives and evaluation of their biological activities.

Deuterium substitution has been widely known that can improve the pharmacokinetic profiles due to isotope effect. Herein, a series of deuterated sorafenib derivatives have been synthesized and characterized by 1H NMR, 13C NMR and MS. Their antitumor activities were evaluated in vitro against human hepatoma cell line HepG2 and human cervical carcinoma cell line HeLa. The LogP values were detected by high-performance liquid chromatography. Subsequently, the metabolic stability and pharmacokinetics study were assessed in vitro and in vivo.

In vitro and in vivo pharmacokinetic and pharmacodynamic study of MBRI-001, a deuterium-substituted plinabulin derivative as a potent anti-cancer agent.

MBRI-001 was demonstrated preliminary better pharmacokinetics and antitumor effects than that of plinabulin in vivo. In this approach, we further carried out systematic pharmacokinetic and pharmacodynamic study of MBRI-001 in vitro and in vivo. MBRI-001 was tested stable in rat plasma and more stable in liver microsomes than plinabulin in vitro. In vivo, MBRI-001 could be distributed rapidly and widely in various tissues, especially the concentration of MBRI-001 in lung was remarkably higher than other tissues. Excretion study indicated that MBRI-001 might been decomposed and excreted as metabolites. Additionally, the combination treatment of MBRI-001 and gefitinib revealed better antitumor inhibition rate than monotherapy in vivo. Therefore, we suggest that MBRI-001 could be developed as a promising anti-cancer agent in near future.

Development of MBRI-001, a deuterium-substituted plinabulin derivative as a potent anti-cancer agent.

Plinabulin, a drug targeting microtubule of cancer cells, has been currently tried in its phase III clinical study. However, low efficacy caused by poor pharmacokinetic (PK) properties has been considered to be the main obstacle to approved by the Food and Drug Administration. Herein, we introduced a deuterium atom as an isostere in its structure to become a new compound named (MBRI-001, No. 9 in a series of deuterium-substituted compounds). The structure of MBRI-001 was characterized by HRMS, NMR, IR and a single crystal analysis. MBRI-001 exhibited better pharmacokinetic characteristics than that of plinabulin. Additionally, its antitumor activity is in a low nanomolar level for a variety of cancer cell lines and high activity against human NCI-H460 xenograted in mice intravenous administration. Importantly, continuous administration of MBRI-001 exhibited lower toxicity compared to docetaxel. We thus suggest that MBRI-001 could be developed as a promising anti-cancer agent in near future.

First total syntheses and antimicrobial evaluation of penicimonoterpene, a marine-derived monoterpenoid, and its various derivatives.

The first total synthesis of marine-derived penicimonoterpene (±)-1 has been achieved in four steps from 6-methylhept-5-en-2-one using a Reformatsky reaction as the key step to construct the basic carbon skeleton. A total of 24 new derivatives of 1 have also been designed and synthesized. Their structures were characterized by analysis of their 1H NMR, 13C NMR and HRESIMS data. Some of them showed significant antibacterial activity against Aeromonas hydrophila, Escherichia coli, Micrococcus luteus, Staphylococcus aureus, Vibrio anguillarum, V. harveyi and/or V. parahaemolyticus, and some showed activity against plant-pathogenic fungi (Alternaria brassicae, Colletotrichum gloeosporioides and/or Fusarium graminearum). Some of the derivatives exhibited antimicrobial MIC values ranging from 0.25 to 4 µg/mL, which were stronger than those of the positive control. Notably, Compounds 3b and 10 showed extremely high selectively against plant-pathogenic fungus F. graminearum (MIC 0.25 µg/mL) and pathogenic bacteria E. coli (MIC 1 µg/mL), implying their potential as antimicrobial agents. SAR analysis of 1 and its derivatives indicated that modification of the carbon-carbon double bond at C-6/7, of groups on the allylic methylene unit and of the carbonyl group at C-1, effectively enhanced the antimicrobial activity.

Health literacy in patients with gout: A latent profile analysis.

OBJECTIVES: Understanding the health literacy status of patients with gout diagnosis is essential for improving the health of this population. Our study aimed to investigate the latent profiles of health literacy in patients with gout and to analyze differences in characteristics across potential profiles. METHODS: This was a cross-sectional study. Eligible participants attended the Shandong Gout Medical Center, from March 2023 to May 2023 and self-reported gout diagnosis. We used the Health Literacy Scale for Patients with Gout designed and validated by our team. The scale had good reliability and validity among patients with gout. 243 patients completed the Demographic Information Questionnaire and the Health Literacy Scale for Patients with Gout. We used latent profile analysis to identify the latent profiles of gout patients' health literacy. We used Chi-square tests with Bonferroni correction to analyze differences in demographics and illness characteristics across identified profiles. RESULTS: Three profiles of patients with gout emerged (prevalence): the low literacy-low critical group (21.81%), the moderate literacy group (42.79%), and the high literacy-stable group (35.39%). The three groups differed in age, education level, monthly income, disease duration, and place of residence (P<0.01). CONCLUSIONS: The health literacy of patients with gout was heterogeneous. Healthcare professionals should adopt targeted interventions based on the characteristics of each latent health literacy profile to improve the health literacy level of patients with gout.

Incidence and risks of subarachnoid hemorrhage in China.

BACKGROUND AND PURPOSE: To determine incidence and risks of subarachnoid hemorrhage in China. METHODS: A prospective, population-based, 1:2 matched case-control study in Baotou, Inner Mongolia (≈2 million population) in 2009-2011. Multiple variable models used to determine relative risk and population-attributable risks for exposures. RESULTS: For a total of 226 patients (mean age, 59 years; 65% women; 434 controls), crude annual incidence (per 100 000) of subarachnoid hemorrhage was 6.2 (95% confidence intervals, 5.4-7.0); 4.3 (3.3-5.2) for men and 8.2 (6.9-9.6) for women. Compared with nonsmokers, adjusted relative risk of subarachnoid hemorrhage in current smokers was 2.31 (95% confidence interval, 1.31-4.09) but was 4.00 (1.62-9.89) in women. Population-attributable risk for smoking, hypertension, and low income were 18%, 36% and 59%, respectively. CONCLUSIONS: The incidence of subarachnoid hemorrhage in China is slightly lower than in Western countries and is related to smoking, hypertension, and poor socioeconomic status.

Benzbromarone, an old uricosuric drug, inhibits human fatty acid binding protein 4 in vitro and lowers the blood glucose level in db/db mice.

AIM: Fatty acid-binding protein 4 (FABP4) plays an important role in maintaining glucose and lipid homeostasis. The aim of this study was to find new inhibitors of FABP4 for the treatment of type 2 diabetes. METHODS: Human FABP4 protein was expressed, and its inhibitors were detected in 1,8-ANS displacement assay. The effect of the inhibitor on lipolysis activity was examined in mouse 3T3-L1 preadipocytes. The db/db mice were used to evaluate the anti-diabetic activity of the inhibitor. Molecular docking and site-directed mutagenesis studies were carried out to explore the binding mode between the inhibitor and FABP4. RESULTS: From 232 compounds tested, benzbromarone (BBR), an old uricosuric drug, was discovered to be the best inhibitor of FABP4 with an IC50 value of 14.8 µmol/L. Furthermore, BBR (25 µmol/L) significantly inhibited forskolin-stimulated lipolysis in 3T3-L1 cells. Oral administration of BBR (25 or 50 mg/kg, for 4 weeks) dose-dependently reduced the blood glucose level and improved glucose tolerance and insulin resistance in db/db mice. Molecular docking revealed that the residues Ser55, Asp76, and Arg126 of FABP4 formed important interactions with BBR, which was confirmed by site-directed mutagenesis studies. CONCLUSION: BBR is an inhibitor of FABP4 and a potential drug candidate for the treatment of type 2 diabetes and atherosclerosis.

Investigating the role of E-commerce marketing capabilities to achieve the strategic performance of tourism firms.

Introduction: This study aims to explore the relationships between strategic performance, e-commerce marketing capabilities (ECMCs), the Internet of Things (IoT) and customer engagement. This study examines the direct association between ECMC and strategic performance. Current research also explores the customer engagement mediation between ECMC and strategic performance (SP). Furthermore, our study investigates the IoT moderation between ECMC and SP. Methods: We test our research hypotheses using data collected in the tourism sector in the context of digital commerce. The questionnaire is used to collect data through random sampling, and these data are useful as a basis for future research. By adding e-commerce capabilities, we show firms how to become more efficient and improve their strategic performance. Moreover, this study, by incorporating the findings of the existing literature, provides a strong foundation for studying the impact of ECMCs and customer engagement on strategic performance as well as the mediating role of e-trust. Results: The results can be useful for managers who conduct digital business internationally, as they need to understand the importance of ECMCs. In fact, the adaptation of ECMCs to the organization enhances customer engagement and helps to improve strategic performance. Discussion: The approach used in this study is in line with previous theoretical analyses and shows emerging patterns in international digital businesses. Moreover, this study adds insights to the e-commerce research by linking different dimensions to reach an in-depth understanding of each item that is affected by ECMCs.

Development and Validation of a Novel UHPLC-MS/MS Method for the Quantification of Plinabulin in Plasma and Its Application in a Pharmacokinetic Study with Leukopenic Rats.

Plinabulin, a new antitumor drug developed from marine natural products that targets microtubules in cancer cells, is currently being tested in a phase III clinical study. Plinabulin has been clinically proven to be effective on leukopenia. However, to our knowledge, there are no reports investigating the pharmacokinetics of plinabulin in individuals with leukopenia and healthy individuals. In this study, we developed a rapid and sensitive UHPLC-MS/MS method for the detection of plinabulin for the first time. Using a novel cyclophosphamide-induced leukopenia model, we investigated the differences in the pharmacokinetic characteristics of plinabulin between rats with leukopenia and normal rats. Plinabulin and propranolol (IS) peaks were separated by gradient elution for a total run time of 5 min. The methodological validation showed a good accuracy (101.96-109.42%) and precision (RSD ≤ 5.37%) with the lower limit of quantification at 0.5 ng/mL. The recovery of plinabulin was between 91.99% and 109.75% (RSD ≤ 7.92%). The values of the area under the plasma concentration-time curve (AUC0-t) for leukopenia groups and control groups at doses of 0.5 mg/kg, 1 mg/kg, and 3 mg/kg were 148.89 ± 78.74 h·µg/L and 121.75 ± 31.56 h·µg/L; 318.15 ± 40.00 h·µg/L and 272.06 ± 42.85 h·µg/L; and 1432.43 ± 197.47 h·µg/L and 1337.12 ± 193.56 h·µg/L; respectively. The half-lives (t1/2s) of plinabulin were 0.49-0.72 h for leukopenia groups and 0.39-0.70 h for control groups at three doses, and the clearance rates (CLs) of plinabulin were 2.13-3.87 L/h/kg for leukopenia groups and 2.29-4.23 L/h/kg for control groups. Pharmacokinetic results showed that there was no significant pharmacokinetic difference between the normal group and the leukopenia group. Based on the power model, plinabulin exhibits a lack of dose proportionality over the dose range of 0.5-3 mg/kg after intravenous administration. This study provides guidance for the development of plinabulin as a potential candidate for the treatment of chemotherapy-induced leukopenia.

Discovery of Novel 2,3-Dihydro-1H-indene-5-sulfonamide NLRP3 Inflammasome Inhibitors Targeting Colon as a Potential Therapy for Colitis.

The NLRP3 inflammasome is a multiprotein complex that plays a crucial role in the pathophysiology of multiple inflammation-related diseases. In this study, we designed and synthesized a series of novel 2,3-dihydro-1H-indene-5-sulfonamide analogues as NLRP3 inflammasome inhibitors, and then identified compound 15z as a potent and specific inhibitor (IC50: 0.13 µM) with low toxicity. Mechanistic studies indicate that 15z binds directly to NLRP3 protein (KD: 102.7 nM), blocking the assembly and activation of the NLRP3 inflammasome and effectively inhibiting cell pyroptosis. Given the notable distribution of 15z in the colon, the DSS-induced colitis model was employed to evaluate its in vivo effectiveness. 15z significantly impacted NLRP3 inflammasome activation and relieved inflammatory bowel disease symptoms in this model. Acute and subacute toxicity studies suggested that 15z has a favorable safety profile. Our results indicate that 15z has great potential to be further developed as a candidate for the treatment of inflammatory bowel disease.

Learning Two-Stream CNN for Multi-Modal Age-Related Macular Degeneration Categorization.

This paper tackles automated categorization of Age-related Macular Degeneration (AMD), a common macular disease among people over 50. Previous research efforts mainly focus on AMD categorization with a single-modal input, let it be a color fundus photograph (CFP) or an OCT B-scan image. By contrast, we consider AMD categorization given a multi-modal input, a direction that is clinically meaningful yet mostly unexplored. Contrary to the prior art that takes a traditional approach of feature extraction plus classifier training that cannot be jointly optimized, we opt for end-to-end multi-modal Convolutional Neural Networks (MM-CNN). Our MM-CNN is instantiated by a two-stream CNN, with spatially-invariant fusion to combine information from the CFP and OCT streams. In order to visually interpret the contribution of the individual modalities to the final prediction, we extend the class activation mapping (CAM) technique to the multi-modal scenario. For effective training of MM-CNN, we develop two data augmentation methods. One is GAN-based CFP/OCT image synthesis, with our novel use of CAMs as conditional input of a high-resolution image-to-image translation GAN. The other method is Loose Pairing, which pairs a CFP image and an OCT image on the basis of their classes instead of eye identities. Experiments on a clinical dataset consisting of 1,094 CFP images and 1,289 OCT images acquired from 1,093 distinct eyes show that the proposed solution obtains better F1 and Accuracy than multiple baselines for multi-modal AMD categorization. Code and data are available at https://github.com/li-xirong/mmc-amd.

Stereoscopic vs. monoscopic photographs on optic disc evaluation and glaucoma diagnosis among general ophthalmologists: A cloud-based real-world multicenter study.

Purpose: To investigate whether stereoscopic vs. monoscopic viewing condition influences the evaluation of optic disc photographs for morphologic features and glaucoma likelihood in a general ophthalmologist population from multicenters on a cloud-based platform. Methods: A cross-sectional study of 519 pairs of stereoscopic and monoscopic photographs of optic discs with adequate quality were selected and presented using a cloud-based platform. A total of 21 general ophthalmologists from 14 centers assessed 15 morphologic features based on 5R's rules and estimated glaucoma likelihood for each assigned photograph. There were 93 pairs of stereoscopic and monoscopic photographs evaluated by a panel of glaucoma specialists and set as ground truth. The main outcome measures were the agreement between estimates and ground truth and the inter-grader agreements. Results: There were good agreements between ground truth and both monoscopic and stereoscopic estimates (stereo κ 0.532 and mono κ 0.494). There was also a substantial intra-grader agreement between monoscopic and stereoscopic evaluation of glaucoma likelihood (κ 0.636). In eyes with probable glaucoma, the accuracy of the stereo method was greater than that of the mono method (stereo 0.238 vs. mono 0.118) When compared with ground truth, stereoscopic photographs had a better agreement for disc size (stereo κ 0.447 vs. mono κ 0.183), disc color (stereo κ 0.612 vs. mono κ 0.549), neuroretinal rim shape (stereo κ 0.356 vs. mono κ 0.274) on the whole. The stereoscopic method also had a better inter-grade agreement for disc size, disc color, neuroretinal rim shape, and glaucoma likelihood (stereo κ 0.402 vs. mono κ 0.359) on the whole. Conclusions: In the evaluation of optic disc photographs for morphologic features and glaucoma likelihood, the stereoscopic method showed superiority compared to the monoscopic method for general ophthalmologists. The stereoscopic method is more likely to identify glaucomatous eyes which need medical intervention.

Three-dimensional diabetic macular edema thickness maps based on fluid segmentation and fovea detection using deep learning.

AIM: To explore a more accurate quantifying diagnosis method of diabetic macular edema (DME) by displaying detailed 3D morphometry beyond the gold-standard quantification indicator-central retinal thickness (CRT) and apply it in follow-up of DME patients. METHODS: Optical coherence tomography (OCT) scans of 229 eyes from 160 patients were collected. We manually annotated cystoid macular edema (CME), subretinal fluid (SRF) and fovea as ground truths. Deep convolution neural networks (DCNNs) were constructed including U-Net, sASPP, HRNetV2-W48, and HRNetV2-W48+Object-Contextual Representation (OCR) for fluid (CME+SRF) segmentation and fovea detection respectively, based on which the thickness maps of CME, SRF and retina were generated and divided by Early Treatment Diabetic Retinopathy Study (ETDRS) grid. RESULTS: In fluid segmentation, with the best DCNN constructed and loss function, the dice similarity coefficients (DSC) of segmentation reached 0.78 (CME), 0.82 (SRF), and 0.95 (retina). In fovea detection, the average deviation between the predicted fovea and the ground truth reached 145.7±117.8 µm. The generated macular edema thickness maps are able to discover center-involved DME by intuitive morphometry and fluid volume, which is ignored by the traditional definition of CRT>250 µm. Thickness maps could also help to discover fluid above or below the fovea center ignored or underestimated by a single OCT B-scan. CONCLUSION: Compared to the traditional unidimensional indicator-CRT, 3D macular edema thickness maps are able to display more intuitive morphometry and detailed statistics of DME, supporting more accurate diagnoses and follow-up of DME patients.

Automated diagnoses of age-related macular degeneration and polypoidal choroidal vasculopathy using bi-modal deep convolutional neural networks.

AIMS: To investigate the efficacy of a bi-modality deep convolutional neural network (DCNN) framework to categorise age-related macular degeneration (AMD) and polypoidal choroidal vasculopathy (PCV) from colour fundus images and optical coherence tomography (OCT) images. METHODS: A retrospective cross-sectional study was proposed of patients with AMD or PCV who came to Peking Union Medical College Hospital. Diagnoses of all patients were confirmed by two retinal experts based on diagnostic gold standard for AMD and PCV. Patients with concurrent retinal vascular diseases were excluded. Colour fundus images and spectral domain OCT images were taken from dilated eyes of patients and healthy controls, and anonymised. All images were pre-labelled into normal, dry or wet AMD or PCV. ResNet-50 models were used as the backbone and alternate machine learning models including random forest classifiers were constructed for further comparison. For human-machine comparison, the same testing data set was diagnosed by three retinal experts independently. All images from the same participant were presented only within a single partition subset. RESULTS: On a test set of 143 fundus and OCT image pairs from 80 eyes (20 eyes per-group), the bi-modal DCNN demonstrated the best performance, with accuracy 87.4%, sensitivity 88.8% and specificity 95.6%, and a perfect agreement with diagnostic gold standard (Cohen's κ 0.828), exceeds slightly over the best expert (Human1, Cohen's κ 0.810). For recognising PCV, the model outperformed the best expert as well. CONCLUSION: A bi-modal DCNN for automated classification of AMD and PCV is accurate and promising in the realm of public health.