HRU-Net: A high-resolution convolutional neural network for esophageal cancer radiotherapy target segmentation.

BACKGROUND AND OBJECTIVE: The effective segmentation of esophageal squamous carcinoma lesions in CT scans is significant for auxiliary diagnosis and treatment. However, accurate lesion segmentation is still a challenging task due to the irregular form of the esophagus and small size, the inconsistency of spatio-temporal structure, and low contrast of esophagus and its peripheral tissues in medical images. The objective of this study is to improve the segmentation effect of esophageal squamous cell carcinoma lesions. METHODS: It is critical for a segmentation network to effectively extract 3D discriminative features to distinguish esophageal cancers from some visually closed adjacent esophageal tissues and organs. In this work, an efficient HRU-Net architecture (High-Resolution U-Net) was exploited for esophageal cancer and esophageal carcinoma segmentation in CT slices. Based on the idea of localization first and segmentation later, the HRU-Net locates the esophageal region before segmentation. In addition, an Resolution Fusion Module (RFM) was designed to integrate the information of adjacent resolution feature maps to obtain strong semantic information, as well as preserve the high-resolution features. RESULTS: Compared with the other five typical methods, the devised HRU-Net is capable of generating superior segmentation results. CONCLUSIONS: Our proposed HRU-NET improves the accuracy of segmentation for squamous esophageal cancer. Compared to other models, our model performs the best. The designed method may improve the efficiency of clinical diagnosis of esophageal squamous cell carcinoma lesions.

Design and Study of a Two-Dimensional (2D) All-Optical Spatial Mapping Module.

Sequentially timed all-optical mapping photography is one of the main emerging ultra-fast detection technologies that can be widely applicable to ultra-fast detection at the picosecond level in fields such as materials and life sciences. We propose a new optical structure for an all-optical spatial mapping module that can control the optical field of two-dimensional imaging while improving spectral resolution and detector sensor utilization. The model of optical parameters based on geometrical optics theory for the given structure has been established, and the theoretical analysis of the inter-frame energy crosstalk caused by incident beam spot width, chromatic aberration, and main errors of the periscope array has been conducted. The optical design of the two-dimensional (2D) all-optical spatial mapping module was finally completed using ZEMAX OpticStudio 2018 software. The results show that our optical module can realize targets of 16 frames and 1.25 nm spectral resolution.

Downregulation of GPX8 in hepatocellular carcinoma: impact on tumor stemness and migration.

PURPOSE: GPX8, which is found in the endoplasmic reticulum lumen, is a member of the Glutathione Peroxidases (GPXs) family. Its role in hepatocellular carcinoma (HCC) is unknown. METHODS: Immunohistochemical staining was used to detect the protein levels of GPX8 in HCC tissue microarrays. A short hairpin RNA lentivirus was used to knock down GPX8, and the main signaling pathways were investigated using transcriptome sequencing and a phosphorylated kinase array. The sphere formation assays, cloning-formation assays and cell migration assays were used to evaluate the stemness and migration ability of HCC cells. Identifying the GPX8-interacting proteins was accomplished through immunoprecipitation and protein mass spectrometry. RESULTS: The GPX8 protein levels were downregulated in HCC patients. Low expression of GPX8 protein was related to early recurrence and poor prognosis in HCC patients. GPX8 knockdown could enhance the stemness and migration ability of HCC cells. Consistently, Based on transcriptome analysis, multiple signaling pathways that include the PI3K-AKT and signaling pathways that regulate the pluripotency of stem cells, were activated after GPX8 knockdown. The downregulation of GPX8 could increase the expression of the tumor stemness markers KLF4, OCT4, and CD133. The in vivo downregulation of GPX8 could also promote the subcutaneous tumor-forming and migration ability of HCC cells. MK-2206, which is a small-molecule inhibitor of AKT, could reverse the tumor-promoting effects both in vivo and in vitro. We discovered that GPX8 and the 71-kDa heat shock cognate protein (Hsc70) have a direct interaction. The phosphorylation of AKT encouraged the translocation of Hsc70 into the nucleus and the expression of the PI3K p110 subunit, thereby increasing the downregulation of GPX8. CONCLUSION: The findings from this study demonstrate the anticancer activity of GPX8 in HCC by inactivating the Hsc70/AKT pathway. The results suggest a possible therapeutic target for HCC.

Spacer Conformation Induced Multiple Hydrogen Bonds in 2D Perovskite toward Highly Efficient Optoelectronic Devices.

Two-dimensional (2D) Dion-Jacobson (DJ) perovskites typically outperform Ruddlesden-Popper (RP) analogs in terms of photodetection (PD). However, the mechanism behind this enhanced performance remains elusive. Theoretical calculations for elucidating interlayer spacer conformation-induced multiple hydrogen bonds in 2D perovskite are presented, along with the synthesis of DPAPbBr4 (DPB) single crystals (SCs) and their PD properties under X-ray/ultraviolet (UV) excitation. The high-quality DPB SC enhances PD with exceptional photoresponse attributes, including a high on/off ratio (4.89 × 104), high responsivity (2.44 A W⁻1), along with large dynamic linear range (154 dB) and low detection limit (7.1 nW cm⁻2), which are currently the best results among 2D perovskite SC detectors, respectively. Importantly, high-resolution images are obtained under UV illumination with weak light levels. The SC X-ray detector exhibits a high sensitivity of 663 µC Gyair⁻1 cm-2 at 10 V and a detection limit of 1.44 µGyair s⁻1. This study explores 2D DJ perovskites for efficient and innovative optoelectronic applications.

A Cross-Sectional and Longitudinal Integrated Study on Brain Functional Changes in a Neuropathic Pain Rat Model.

Human and animal imaging studies demonstrated that chronic pain profoundly alters the structure and the functionality of several brain regions and even causes mental dysfunctions such as depression and anxiety disorders. In this article, we conducted a multimodal study cross-sectionally and longitudinally, to evaluate how neuropathic pain affects the brain. Using the spared nerve injury (SNI) model which promotes long-lasting mechanical allodynia, results showed that neuropathic pain deeply modified the intrinsic organization of the brain functional network 2 weeks after injury. There are significant changes in the activity of the left thalamus (Th_L) and left olfactory bulb (OB_L) brain regions after SNI, as evidenced by both the blood oxygen level-dependent (BOLD) signal and c-Fos expression. Importantly, these changes were closely related to mechanical pain behavior of rats. However, it is worth noting that after morphine administration for analgesia, only the increased activity in the TH region is reversed, while the decreased activity in the OB region becomes more prominent. Functional connectivity (FC) and c-Fos correlation analysis further showed these two regions of interest (ROIs) exhibit different FC patterns with other brain regions. Our study comprehensively revealed the adaptive changes of brain neural networks induced by nerve injury in both cross-sectional and longitudinal dimensions and emphasized the abnormal activity and FC of Th_L and OB_L in the pathological condition. It provides reliable assistance in exploring the intricate mechanisms of diseases.

Association of height, foot length, and pubertal development in children aged 3-18: a cross-sectional survey in China.

Objective: This study aimed to analyze the growth patterns of height and foot length (FL) among Chinese children aged 3-18 and examine their associations with puberty development. Methods: A cross-sectional survey was conducted in September 2022 in Beijing. Data were collected through questionnaires and on-site physical examinations. The growth patterns and velocity of height and FL in different age groups were described, and their associations with puberty development were analyzed. Results: From an age perspective, the peak FL growth occurred between 9 and 11 years (boys were 11 years and girls were 9 years), while the peak height growth occurred at 11 ~ 13 years for boys and 9 ~ 11 years for girls. Additionally, boys and girls reached 99.0% of their final FL at the ages of 14 and 13, respectively, while they reached 99.0% of their final height at the ages of 16 and 15, respectively. From the perspective of Tanner stage, the age of peak FL growth in boys coincided with the age of the G2 stage, while in girls it occurred slightly earlier than the mean age of the B2 stage. The peak height growth for both boys and girls occurred between Tanner stages 2 and 3. Conclusion: Boys and girls reach their peak FL growth at 11 and 9 years old, respectively, which were both 2 years earlier than their peak height growth. The peak FL growth occurred around the onset of puberty, while the peak height growth occurred between Tanner stages 2 and 3.

Manipulation of Electronic States of Pt Sites via d-Band Center Tuning for Enhanced Oxygen Reduction Reaction in Proton Exchange Membrane Fuel Cells.

Expediting the torpid kinetics of the oxygen reduction reaction (ORR) at the cathode with minimal amounts of Pt under acidic conditions plays a significant role in the development of proton exchange membrane fuel cells (PEMFCs). Herein, a novel Pt-N-C system consisting of Pt single atoms and nanoparticles anchored onto the defective carbon nanofibers is proposed as a highly active ORR catalyst (denoted as Pt-N-C). Detailed characterizations together with theoretical simulations illustrate that the strong coupling effect between different Pt sites can enrich the electron density of Pt sites, modify the d-band electronic environments, and optimize the oxygen intermediate adsorption energies, ultimately leading to significantly enhanced ORR performance. Specifically, the as-designed Pt-N-C demonstrates exceptional ORR properties with a high half-wave potential of 0.84 V. Moreover, the mass activity of Pt-N-C reaches 193.8 mA gPt-1 at 0.9 V versus RHE, which is 8-fold greater than that of Pt/C, highlighting the enormously improved electrochemical properties. More impressively, when integrated into a membrane electrode assembly as cathode in an air-fed PEMFC, Pt-N-C achieved a higher maximum power density (655.1 mW cm-2) as compared to Pt/C-based batteries (376.25 mW cm-2), hinting at the practical application of Pt-N-C in PEMFCs.

Single-cell and spatial architecture of primary liver cancer.

Primary liver cancer (PLC) poses a leading threat to human health, and its treatment options are limited. Meanwhile, the investigation of homogeneity and heterogeneity among PLCs remains challenging. Here, using single-cell RNA sequencing, spatial transcriptomic and bulk multi-omics, we elaborated a molecular architecture of 3 PLC types, namely hepatocellular carcinoma (HCC), intrahepatic cholangiocarcinoma (ICC) and combined hepatocellular-cholangiocarcinoma (CHC). Taking a high-resolution perspective, our observations revealed that CHC cells exhibit internally discordant phenotypes, whereas ICC and HCC exhibit distinct tumor-specific features. Specifically, ICC was found to be the primary source of cancer-associated fibroblasts, while HCC exhibited disrupted metabolism and greater individual heterogeneity of T cells. We further revealed a diversity of intermediate-state cells residing in the tumor-peritumor junctional zone, including a congregation of CPE+ intermediate-state endothelial cells (ECs), which harbored the molecular characteristics of tumor-associated ECs and normal ECs. This architecture offers insights into molecular characteristics of PLC microenvironment, and hints that the tumor-peritumor junctional zone could serve as a targeted region for precise therapeutical strategies.

Adolescent pursuit of health information online during the COVID-19 pandemic: The roles played by eHealth literacy and psychological distress.

COVID-19 has led to an increase in mental health problems for adolescents. In this study, we examined the factors related to the eHealth literacy of adolescents and how that impacted their pursuit of health information and mental health information online during the COVID-19 pandemic. We analyzed data from the 2020 Taiwan Communication Survey, which involved a total of 1,250 national representative adolescents who completed an online questionnaire. The results showed that two-thirds of adolescents reported searching for health information online, and about half of them reported searching for mental health information online during the COVID-19 pandemic. Multivariate analysis results indicated that adolescents who spent more time learning online, had higher levels of bonding social capital and self-determination, and had higher levels of parental active internet mediation were more likely to have higher levels of eHealth literacy. In addition, multivariate analysis results showed that adolescents who had higher levels of eHealth literacy and had higher depression and anxiety were more likely to seek health information and mental health information online. In conclusion, the levels of eHealth literacy and psychological distress of adolescents played a crucial role in their pursuit of health information and mental health information online during the COVID-19 pandemic.

Halide Ions Regulating the Morphologies of PbS and Au@PbS Core-Shell Nanocrystals: Synthesis, Self-Assembly, and Electrical Transport Properties.

The geometry and surface state of nanocrystals (NCs) strongly affect their physiochemical properties, self-assembly behaviors, and potential applications, but there is still a lack of a facile method to regulate the exposed facets of the NCs, especially metal@semiconductor core-shell NCs. Herein, we present a reproducible approach for tuning the morphology of PbS NCs from nanocubes to nano-octahedrons by introducing lead halides as precursors. Excitingly, the method can be easily extended to the synthesis of metal@PbS core-shell NCs with single-crystalline shells and specific exposed facets. In addition, the halide passivation layers on the NCs are found to greatly improve their antioxidant ability. Therefore, the Cl-passivated NCs can self-assemble into atomic-coupled monolayer films via oriented attachment under ambient conditions, which exhibit enhanced electrical conductivities compared with uncoupled counterparts. The precise synthesis of nanocrystals with tunable shapes and the construction of self-assembled films provide a way to expand their application in high-performance optoelectronic devices.

Biological and clinical implications of early-onset cancers: A unique subtype.

In recent years, the incidence of cancers is continuously increasing in young adults. Early-onset cancer (EOC) is usually defined as patients with cancers under the age of 50, and may represent a unique subgroup due to its special disease features. Overall, EOCs often initiate at a young age, present as a better physical performance but high degree of malignancy. EOCs also share common epidemiological and hereditary risk factors. In this review, we discuss several representative EOCs which were well studied previously. By revealing their clinical and molecular similarities and differences, we consider the group of EOCs as a unique subtype compared to ordinary cancers. In consideration of EOC as a rising threat to human health, more researches on molecular mechanisms, and large-scale, prospective clinical trials should be carried out to further translate into improved outcomes.

Versatile synthesis of nano-icosapods via cation exchange for effective photocatalytic conversion of biomass-relevant alcohols.

Branched metal chalcogenide nanostructures with well-defined composition and configuration are appealing photocatalysts for solar-driven organic transformations. However, precise design and controlled synthesis of such nanostructures still remain a great challenge. Herein, we report the construction of a variety of highly symmetrical metal sulfides and heterostructured icosapods based on them, in which twenty branches were radially grown in spatially ordered arrangement, with a high degree of structure homogeneity. Impressively, the as-obtained CdS-PdxS icosapods manifest a significantly improved photocatalytic activity for the selective oxidation of biomass-relevant alcohols into corresponding aldehydes coupled with H2 evolution under visible-light irradiation (>420 nm), and the apparent quantum yield of the benzyl alcohol reforming can be achieved as high as 31.4% at 420 nm. The photoreforming process over the CdS-PdxS icosapods is found to be directly triggered by the photogenerated electrons and holes without participation of radicals. The enhanced photocatalytic performance is attributed to the fast charge separation and abundant active sites originating from the well-defined configuration and spatial organization of the components in the branched heterostructures.

Surface Reconstruction and Passivation of BiVO4 Photoanodes Depending on the "Structure Breaker" Cs.

Monoclinic BiVO4 is one of the most promising photoanode materials for solar water splitting. The photoelectrochemical performance of a BiVO4 photoanode could be significantly influenced by the noncovalent interactions of redox-inert metal cations at the photoanode-electrolyte interfaces, but this point has not been well investigated. In this work, we studied the Cs+-dependent surface reconstruction and passivation of BiVO4 photoanodes. Owing to the "structure breaker" nature of Cs+, the Cs+ at the BiVO4 photoanode-electrolyte interfaces participated in BiVO4 surface photocorrosion to form a Cs+-doped bismuth vanadium oxide amorphous thin layer, which inhibited the continuous photocorrosion of BiVO4 and promoted surface charge transfer and water oxidation. The resulting cocatalyst-free BiVO4 photoanodes achieved 3.3 mA cm-2 photocurrent for water oxidation. With the modification of FeOOH catalysts, the photocurrent at 1.23 VRHE reached 5.1 mA cm-2, and a steady photocurrent of 3.0 mA cm-2 at 0.8 VRHE was maintained for 30 h. This work provides new insights into the understanding of Cs+ chemistry and the effects of redox-inert cations at the electrode-electrolyte interfaces.

Epitaxial growth of highly symmetrical branched noble metal-semiconductor heterostructures with efficient plasmon-induced hot-electron transfer.

Epitaxial growth is one of the most commonly used strategies to precisely tailor heterostructures with well-defined compositions, morphologies, crystal phases, and interfaces for various applications. However, as epitaxial growth requires a small interfacial lattice mismatch between the components, it remains a challenge for the epitaxial synthesis of heterostructures constructed by materials with large lattice mismatch and/or different chemical bonding, especially the noble metal-semiconductor heterostructures. Here, we develop a noble metal-seeded epitaxial growth strategy to prepare highly symmetrical noble metal-semiconductor branched heterostructures with desired spatial configurations, i.e., twenty CdS (or CdSe) nanorods epitaxially grown on twenty exposed (111) facets of Ag icosahedral nanocrystal, albeit a large lattice mismatch (more than 40%). Importantly, a high quantum yield (QY) of plasmon-induced hot-electron transferred from Ag to CdS was observed in epitaxial Ag-CdS icosapods (18.1%). This work demonstrates that epitaxial growth can be achieved in heterostructures composed of materials with large lattice mismatches. The constructed epitaxial noble metal-semiconductor interfaces could be an ideal platform for investigating the role of interfaces in various physicochemical processes.

Development of a deep pathomics score for predicting hepatocellular carcinoma recurrence after liver transplantation.

BACKGROUND AND PURPOSE: Tumor recurrence after liver transplantation (LT) impedes the curative chance for hepatocellular carcinoma (HCC) patients. This study aimed to develop a deep pathomics score (DPS) for predicting tumor recurrence after liver transplantation using deep learning. PATIENTS AND METHODS: Two datasets of 380 HCC patients who underwent LT were enrolled. Residual convolutional neural networks were used to identify six histological structures of HCC. The individual risk score of each structure and DPS were derived by a modified DeepSurv network. Cox regression analysis and Concordance index were used to evaluate the prognostic significance. The cellular exploration of prognostic immune biomarkers was performed by quantitative and spatial proximity analysis according to three panels of 7-color immunofluorescence. RESULTS: The overall classification accuracy of HCC tissue was 97%. At the structural level, immune cells were the most significant tissue category for predicting post-LT recurrence (HR 1.907, 95% CI 1.490-2.440). The C-indices of DPS achieved 0.827 and 0.794 in the training and validation cohorts, respectively. Multivariate analysis for recurrence-free survival (RFS) showed that DPS (HR 4.795, 95% CI 3.017-7.619) was an independent risk factor. Patients in the high-risk subgroup had a shorter RFS, larger tumor diameter and a lower proportion of clear tumor borders. At the cellular level, a higher infiltration of intratumoral NK cells was negatively correlated with recurrence risk. CONCLUSIONS: This study established an effective DPS. Immune cells were the most significant histological structure related to HCC recurrence. DPS performed well in post-LT recurrence prediction and the identification of clinicopathological features.

Antimony Potassium Tartrate Stabilizes Wide-Bandgap Perovskites for Inverted 4-T All-Perovskite Tandem Solar Cells with Efficiencies over 26.

Wide-bandgap (WBG) perovskites have been attracting much attention because of their immense potential as a front light-absorber for tandem solar cells. However, WBG perovskite solar cells (PSCs) generally exhibit undesired large open-circuit voltage (VOC) loss due to light-induced phase segregation and severe non-radiative recombination loss. Herein, antimony potassium tartrate (APTA) is added to perovskite precursor as a multifunctional additive that not only coordinates with unbonded lead but also inhibits the migration of halogen in perovskite, which results in suppressed non-radiative recombination, inhibited phase segregation and better band energy alignment. Therefore, a APTA auxiliary WBG PSC with a champion photoelectric conversion efficiency of 20.35% and less hysteresis is presented. They maintain 80% of their initial efficiencies under 100 mW cm-2 white light illumination in nitrogen after 1,000 h. Furthermore, by combining a semi-transparent WBG perovskite front cell with a narrow-bandgap tin-lead PSC, a perovskite/perovskite four-terminal tandem solar cell with an efficiency over 26% is achieved. Our work provides a feasible approach for the fabrication of efficient tandem solar cells.

Clinical and biomarker analyses of sintilimab plus gemcitabine and cisplatin as first-line treatment for patients with advanced biliary tract cancer.

The prognosis of biliary tract cancer (BTC) remains unsatisfactory. This single-arm, phase II clinical trial (ChiCTR2000036652) investigated the efficacy, safety, and predictive biomarkers of sintilimab plus gemcitabine and cisplatin as the first-line treatment for patients with advanced BTCs. The primary endpoint was overall survival (OS). Secondary endpoints included toxicities, progression-free survival (PFS), and objective response rate (ORR); multi-omics biomarkers were assessed as exploratory objective. Thirty patients were enrolled and received treatment, the median OS and PFS were 15.9 months and 5.1 months, the ORR was 36.7%. The most common grade 3 or 4 treatment-related adverse events were thrombocytopenia (33.3%), with no reported deaths nor unexpected safety events. Predefined biomarker analysis indicated that patients with homologous recombination repair pathway gene alterations or loss-of-function mutations in chromatin remodeling genes presented better tumor response and survival outcomes. Furthermore, transcriptome analysis revealed a markedly longer PFS and tumor response were associated with higher expression of a 3-gene effector T cell signature or an 18-gene inflamed T cell signature. Sintilimab plus gemcitabine and cisplatin meets pre-specified endpoints and displays acceptable safety profile, multiomics potential predictive biomarkers are identified and warrant further verification.

Intermediate Phase Engineering with 2,2-Azodi(2-Methylbutyronitrile) for Efficient and Stable Perovskite Solar Cells.

Sequential deposition has been widely employed to modulate the crystallization of perovskite solar cells because it can avoid the formation of nucleation centers and even initial crystallization in the precursor solution. However, challenges remain in overcoming the incomplete and random transformation of PbI2 films with organic ammonium salts. Herein, a unique intermediate phase engineering strategy has been developed by simultaneously introducing 2,2-azodi(2-methylbutyronitrile) (AMBN) to both PbI2 and ammonium salt solutions to regulate perovskite crystallization. AMBN not only coordinates with PbI2 to form a favorably mesoporous PbI2 film due to the coordination between Pb2+ and the cyano group (C≡N), but also suppresses the vigorous activity of FA+ ions by interacting with FAI, leading to the full PbI2 transformation with the preferred orientation. Therefore, perovskites with favorable facet orientations are obtained, and the defects are largely suppressed owing to the passivation of uncoordinated Pb2+ and FA+ . As a result, a champion power conversion efficiency over 25% with a stabilized efficiency of 24.8% is achieved. Moreover, the device exhibits an improved operational stability, retaining 96% of initial power conversion efficiency under 1000 h continuous white-light illumination with an intensity of 100 mW cm-2 at ≈55 °C in N2 atmosphere.

CD36+ cancer-associated fibroblasts provide immunosuppressive microenvironment for hepatocellular carcinoma via secretion of macrophage migration inhibitory factor.

Hepatocellular carcinoma (HCC) is an immunotherapy-resistant malignancy characterized by high cellular heterogeneity. The diversity of cell types and the interplay between tumor and non-tumor cells remain to be clarified. Single cell RNA sequencing of human and mouse HCC tumors revealed heterogeneity of cancer-associated fibroblast (CAF). Cross-species analysis determined the prominent CD36+ CAFs exhibited high-level lipid metabolism and expression of macrophage migration inhibitory factor (MIF). Lineage-tracing assays showed CD36+CAFs were derived from hepatic stellate cells. Furthermore, CD36 mediated oxidized LDL uptake-dependent MIF expression via lipid peroxidation/p38/CEBPs axis in CD36+ CAFs, which recruited CD33+myeloid-derived suppressor cells (MDSCs) in MIF- and CD74-dependent manner. Co-implantation of CD36+ CAFs with HCC cells promotes HCC progression in vivo. Finally, CD36 inhibitor synergizes with anti-PD-1 immunotherapy by restoring antitumor T-cell responses in HCC. Our work underscores the importance of elucidating the function of specific CAF subset in understanding the interplay between the tumor microenvironment and immune system.

Triple-DRNet: A triple-cascade convolution neural network for diabetic retinopathy grading using fundus images.

Diabetic Retinopathy (DR) is a universal ocular complication of diabetes patients and also the main disease that causes blindness in the world wide. Automatic and efficient DR grading acts a vital role in timely treatment. However, it is difficult to effectively distinguish different types of distinct lesions (such as neovascularization in proliferative DR, microaneurysms in mild NPDR, etc.) using traditional convolutional neural networks (CNN), which greatly affects the ultimate classification results. In this article, we propose a triple-cascade network model (Triple-DRNet) to solve the aforementioned issue. The Triple-DRNet effectively subdivides the classification of five types of DR as well as improves the grading performance which mainly includes the following aspects: (1) In the first stage, the network carries out two types of classification, namely DR and No DR. (2) In the second stage, the cascade network is intended to distinguish the two categories between PDR and NPDR. (3) The final cascade network will be designed to differentiate the mild, moderate and severe types in NPDR. Experimental results show that the ACC of the Triple-DRNet on the APTOS 2019 Blindness Detection dataset achieves 92.08%, and the QWK metric reaches 93.62%, which proves the effectiveness of the devised Triple-DRNet compared with other mainstream models.