The performance of a deep learning system in assisting junior ophthalmologists in diagnosing 13 major fundus diseases: a prospective multi-center clinical trial.

Artificial intelligence (AI)-based diagnostic systems have been reported to improve fundus disease screening in previous studies. This multicenter prospective self-controlled clinical trial aims to evaluate the diagnostic performance of a deep learning system (DLS) in assisting junior ophthalmologists in detecting 13 major fundus diseases. A total of 1493 fundus images from 748 patients were prospectively collected from five tertiary hospitals in China. Nine junior ophthalmologists were trained and annotated the images with or without the suggestions proposed by the DLS. The diagnostic performance was evaluated among three groups: DLS-assisted junior ophthalmologist group (test group), junior ophthalmologist group (control group) and DLS group. The diagnostic consistency was 84.9% (95%CI, 83.0% ~ 86.9%), 72.9% (95%CI, 70.3% ~ 75.6%) and 85.5% (95%CI, 83.5% ~ 87.4%) in the test group, control group and DLS group, respectively. With the help of the proposed DLS, the diagnostic consistency of junior ophthalmologists improved by approximately 12% (95% CI, 9.1% ~ 14.9%) with statistical significance (P < 0.001). For the detection of 13 diseases, the test group achieved significant higher sensitivities (72.2% ~ 100.0%) and comparable specificities (90.8% ~ 98.7%) comparing with the control group (sensitivities, 50% ~ 100%; specificities 96.7 ~ 99.8%). The DLS group presented similar performance to the test group in the detection of any fundus abnormality (sensitivity, 95.7%; specificity, 87.2%) and each of the 13 diseases (sensitivity, 83.3% ~ 100.0%; specificity, 89.0 ~ 98.0%). The proposed DLS provided a novel approach for the automatic detection of 13 major fundus diseases with high diagnostic consistency and assisted to improve the performance of junior ophthalmologists, resulting especially in reducing the risk of missed diagnoses. ClinicalTrials.gov NCT04723160.

A highly sensitive and anti-freezing conductive strain sensor based on polypyrrole/cellulose nanofiber crosslinked polyvinyl alcohol hydrogel for human motion detection.

Electro-conductive hydrogels emerge as a stretchable conductive materials with diverse applications in the synthesis of flexible strain sensors. However, the high-water content and low cross-links density cause them to be mechanically destroyed and freeze at subzero temperatures, limiting their practical applications. Herein, we report a one-pot strategy by co-incorporating cellulose nanofiber (CNF), Poly pyrrole (PPy) and glycerol with polyvinyl alcohol (PVA) to prepare hydrogel. The addition of PPy endowed the hydrogel with good conductivity (∼0.034 S/m) compared to the no PPy@CNF group (∼0.0095 S/m), the conductivity was increased by 257.9 %. The hydrogel exhibits comparable ionic conductivity at -18 °C as it does at room temperature. It's attributed to the glycerol as a cryoprotectant and the formation of hydrated [Zn(H2O)n]2+ ions via strong interaction between Zn2+ and water molecules. Moreover, the cellulose nanofiber intrinsically assembled into unique hierarchical structures allow for strong hydrogen bonds between adjacent cellulose and PPy polymer chains, greatly improve the mechanical strength (stress∼0.65 MPa, strain∼301 %) and excellent viscoelasticity (G'max âˆ¼ 82.7 KPa). This novel PPy@CNF-PVA hydrogel exhibits extremely high Gauge factor (GF) of 2.84 and shows excellent sensitivity, repeatability and stability. Therefore, the hydrogel can serve as reliable and stable strain sensor which shows excellent responsiveness in human activities monitoration.

Retinopathy is associated with impaired cognition in patients undergoing peritoneal dialysis.

Objective: Previous studies have shown a relationship between retinopathy and cognition including population with and without chronic kidney disease (CKD) but data regarding peritoneal dialysis (PD) are limited. This study aims to investigate the relationship between retinopathy and cognitive impairment in patients undergoing peritoneal dialysis (PD). Methods: In this observational study, we recruited a total of 107 participants undergoing PD, consisting of 48 men and 59 women, ages ranging from 21 to 78 years. The study followed a cross-sectional design. Retinal microvascular characteristics, such as geometric changes in retinal vascular including tortuosity, fractal dimension (FD), and calibers, were assessed. Retinopathy (such as retinal hemorrhage or microaneurysms) was evaluated using digitized photographs. The Modified Mini-Mental State Examination (3MS) was performed to assess global cognitive function. Results: The prevalence rates of retinal hemorrhage, microaneurysms, and retinopathy were 25%, 30%, and 43%, respectively. The mean arteriolar and venular calibers were 63.2 and 78.5 µm, respectively, and the corresponding mean tortuosity was 37.7 ± 3.6 and 37.2 ± 3.0 mm-1. The mean FD was 1.49. After adjusting for age, sex, education, mean arterial pressure, and Charlson index, a negative association was revealed between retinopathy and 3MS scores (regression coefficient: -3.71, 95% confidence interval: -7.09 to -0.33, p = 0.03). Conclusions: Retinopathy, a condition common in patients undergoing PD, was associated with global cognitive impairment. These findings highlight retinopathy, can serve as a valuable primary screening tool for assessing the risk of cognitive decline.

Multi-physics coupling reinforced polyvinyl alcohol/cellulose nanofibrils based multifunctional hydrogel sensor for human motion monitoring.

Ionic conductive hydrogels have been widely used for sensor, energy storage and human-machine interface. To address the problems of the traditional ionic conductive hydrogels fabricated with the soaking method, such as the lack of frost resistance, poor mechanical properties, time-consuming and chemical-wasting, herein, a multi-physics crosslinking reinforced strong, anti-freezing and ionic conductive hydrogel sensor is fabricated utilizing the tannin acid-Fe2(SO4)3 through the simple one-pot freezing-thawing process at low electrolyte concentration. The results show that the P10C0.4T8-Fe2(SO4)3 (PVA10%CNF0.4%TA8%-Fe2(SO4)3) displayed better mechanical property and ionic conductivity due to hydrogen bonding and coordination interaction. The tensile stress reaches up to 0.980 MPa (570 % strain). Moreover, the hydrogel presents excellent ionic conductivity (0.220 S⋅m-1 at room temperature), anti-freezing performance (0.183 S⋅m-1 at -18 °C), large gauge factor (1.75), excellent sensing stability, repeatability, durability and reliability. This work paves a way for preparing mechanical strong and anti-freezing hydrogel based on multi-physics crosslinking with one-pot freezing-thawing process.

Quercetin Mediated TET1 Expression Through MicroRNA-17 Induced Cell Apoptosis in Melanoma Cells.

A previous report suggested that the expression of ten-eleven translocation (TET) proteins is abnormal in certain cancers. Quercetin has been demonstrated as anti-cancer role in cancer development. In order to explore the inhibitory effect and mechanism of quercetin on uveal melanoma cells, the expression of TET proteins was analyzed in the present study. Our results suggest that the expression of TET1 was increased following treatment with quercetin in OCM-1, SK-MEL-1, and B16 cells. In addition, quercetin treatment induced apoptosis and inhibited migration and invasion. To further investigate the association of the expression of TET1 with cell growth, apoptosis, migration, and invasion, cell lines in which TET1 was knocked-down or overexpressed were constructed. The results showed that the increased expression of TET1-induced apoptosis, increased 5-hydroxymethylcytosine (5 hmC). and inhibited invasion. Our bioinformatics studies indicated that TET1 is a target gene of microRNA-17 (miR-17) Our results showed that inhibition of the expression of miR-17 resulted in increased TET1 expression in OCM-1 cells. Furthermore, our results indicated that quercetin treatment increased TET1 expression and inhibited melanoma growth in nude mice. Taken together, our results suggest that quercetin can regulate cell proliferation and apoptosis through TET1 via miR-17 in melanoma cells.

Prognostic value of antitumor drug targets prediction using integrated bioinformatic analysis for immunogenic cell death-related lncRNA model based on stomach adenocarcinoma characteristics and tumor immune microenvironment.

Stomach adenocarcinoma (STAD) ranks as the fourth prevalent cause of mortality worldwide due to cancer. The prognosis for those suffering from STAD was bleak. Immunogenic cell death (ICD), a form of induced cellular death that causes an adaptive immune response and has increasing in anticancer treatment. However, it has not been ascertained how ICD-related lncRNAs affect STAD. Using univariate Cox regression and the TCGA database, lncRNAs with prognostic value were identified. Thereafter, we created a prognostic lncRNA-based model using LASSO. Kaplan-Meier assessment, time-dependent receiver operating characteristic (ROC) analyzation, independent prognostic investigation, and nomogram were used to assess model correctness. Additional research included evaluations of the immunological microenvironment, gene set enrichment analyses (GSEA), tumor mutation burdens (TMBs), tumor immune dysfunctions and exclusions (TIDEs), and antitumor compounds IC50 predictions. We found 24 ICD-related lncRNAs with prognostic value via univariate Cox analysis (p < 0.05). Subsequently, a risk model was proposed using five lncRNAs relevant to ICD. The risk signature, correlated with immune cell infiltration, had strong predictive performance. Individuals at low-risk group outlived those at high risk (p < 0.001). An evaluation of the 5-lncRNA risk mode including ROC curves, nomograms, and correction curves confirmed its predictive capability. The findings of functional tests revealed a substantial alteration in immunological conditions and the IC50 sensitivity for the two groups. Using five ICD-related lncRNAs, the authors developed a new risk model for STAD patients that could predict their cumulative overall survival rate and guide their individual treatment.

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.

Screening of biomarkers associated with diagnosis and prognosis of colorectal cancer.

We aimed to explore biomarkers associated with diagnosis and prognosis of colorectal cancer. Differentially expressed protein (DEP) genes were obtained and validated. Moreover, co-expressed genes were screened and their prognostic value was evaluated. In addition, miRNAs that were negatively correlated with DEP genes were identified and used to construct a competitive endogenous RNA network. Furthermore, a support vector machine model was built using DEP genes, and a receiver operating characteristic curve was implemented to confirm its prediction performance. The results showed that only one DEP gene, CCL26, was obtained. Moreover, 43 genes co-expressed with CCL26 were identified, among which six (AP3M2, DAPK1, ISYNA1, PPM1K, PRR4 and RNF122) were linked with the prognosis of colorectal cancer. Besides, the axis RP11-47122.2/RP11-527N22.1-hsa-miR-3192-5p-CCL26 was identified as an lncRNA-miRNA-target gene network. Support vector machine model analysis showed that the area under the curve of CCL26 reached 0.878 based on GEO data and 0.743 based on our protein data. In conclusion, AP3M2, DAPK1, ISYNA1, PPM1K, PRR4, RNF122, CCL26 and hsa-miR-3192-5p appear to be related to the progression of colorectal cancer.

An Artificial-Intelligence-Based Automated Grading and Lesions Segmentation System for Myopic Maculopathy Based on Color Fundus Photographs.

Purpose: To develop deep learning models based on color fundus photographs that can automatically grade myopic maculopathy, diagnose pathologic myopia, and identify and segment myopia-related lesions. Methods: Photographs were graded and annotated by four ophthalmologists and were then divided into a high-consistency subgroup or a low-consistency subgroup according to the consistency between the results of the graders. ResNet-50 network was used to develop the classification model, and DeepLabv3+ network was used to develop the segmentation model for lesion identification. The two models were then combined to develop the classification-and-segmentation-based co-decision model. Results: This study included 1395 color fundus photographs from 895 patients. The grading accuracy of the co-decision model was 0.9370, and the quadratic-weighted κ coefficient was 0.9651; the co-decision model achieved an area under the receiver operating characteristic curve of 0.9980 in diagnosing pathologic myopia. The photograph-level F1 values of the segmentation model identifying optic disc, peripapillary atrophy, diffuse atrophy, patchy atrophy, and macular atrophy were all >0.95; the pixel-level F1 values for segmenting optic disc and peripapillary atrophy were both >0.9; the pixel-level F1 values for segmenting diffuse atrophy, patchy atrophy, and macular atrophy were all >0.8; and the photograph-level recall/sensitivity for detecting lacquer cracks was 0.9230. Conclusions: The models could accurately and automatically grade myopic maculopathy, diagnose pathologic myopia, and identify and monitor progression of the lesions. Translational Relevance: The models can potentially help with the diagnosis, screening, and follow-up for pathologic myopic in clinical practice.

The Validation of Deep Learning-Based Grading Model for Diabetic Retinopathy.

Purpose: To evaluate the performance of a deep learning (DL)-based artificial intelligence (AI) hierarchical diagnosis software, EyeWisdom V1 for diabetic retinopathy (DR). Materials and Methods: The prospective study was a multicenter, double-blind, and self-controlled clinical trial. Non-dilated posterior pole fundus images were evaluated by ophthalmologists and EyeWisdom V1, respectively. The diagnosis of manual grading was considered as the gold standard. Primary evaluation index (sensitivity and specificity) and secondary evaluation index like positive predictive values (PPV), negative predictive values (NPV), etc., were calculated to evaluate the performance of EyeWisdom V1. Results: A total of 1,089 fundus images from 630 patients were included, with a mean age of (56.52 ± 11.13) years. For any DR, the sensitivity, specificity, PPV, and NPV were 98.23% (95% CI 96.93-99.08%), 74.45% (95% CI 69.95-78.60%), 86.38% (95% CI 83.76-88.72%), and 96.23% (95% CI 93.50-98.04%), respectively; For sight-threatening DR (STDR, severe non-proliferative DR or worse), the above indicators were 80.47% (95% CI 75.07-85.14%), 97.96% (95% CI 96.75-98.81%), 92.38% (95% CI 88.07-95.50%), and 94.23% (95% CI 92.46-95.68%); For referral DR (moderate non-proliferative DR or worse), the sensitivity and specificity were 92.96% (95% CI 90.66-94.84%) and 93.32% (95% CI 90.65-95.42%), with the PPV of 94.93% (95% CI 92.89-96.53%) and the NPV of 90.78% (95% CI 87.81-93.22%). The kappa score of EyeWisdom V1 was 0.860 (0.827-0.890) with the AUC of 0.958 for referral DR. Conclusion: The EyeWisdom V1 could provide reliable DR grading and referral recommendation based on the fundus images of diabetics.

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.

Porous Carbon Material Derived from Steam-Exploded Poplar for Supercapacitor: Insights into Synergistic Effect of KOH and Urea on the Structure and Electrochemical Properties.

The electrochemical performance of supercapacitors using porous carbon as electrodes is strongly affected by the fabrication process of carbon material. KOH is commonly used as an activator combined with urea as a nitrogen dopant. However, the roles of KOH and urea in pore structure configuration and the electrochemical behavior of porous carbon electrodes are still ambiguous. Herein, the optimum porous carbon is obtained when KOH and urea are used simultaneously. KOH is used as a pore-forming substance, whereas urea is employed as a nitrogen source for the nitrogen doping of porous carbon, which increases its defect sites while reducing the graphitization degree. More importantly, urea also expands pores as a pore-enlarging agent, inducing interconnected porous structures. As a result, a hierarchical porous structure is formed and ascribed to the synergistic effect of KOH and urea, and the specific surface area reached 3282 m2 g-1 for sample PC800-4. The specific capacitance is 319 F g-1 at 0.5 A g-1 with excellent cycling stability over 2500 cycles. Furthermore, the symmetric supercapacitor reaches an excellent energy density of 11.6 W h kg-1 under 70.0 W kg-1 in a 6 M KOH electrolyte. Our work contributes to the rational designation of the porous carbon structure for supercapacitor applications.

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.

One-pot freezing-thawing preparation of cellulose nanofibrils reinforced polyvinyl alcohol based ionic hydrogel strain sensor for human motion monitoring.

Ionic conductive hydrogels have been widely applied in sensors, energy storage and soft electronics recently. However, most of the polyvinyl alcohol (PVA) based ionic hydrogels are mainly fabricated by soaking the hydrogels in high concentration electrolyte solution which can induce the waste of electrolyte and solvent. Herein, we have designed cellulose nanofibrils (CNF) and ZnSO4 reinforced PVA based hydrogels through a one-pot simple freezing-thawing method at low ZnSO4 concentration without any soaking process. Furthermore, the hydrogel with 0.4% CNF exhibited stress up to 0.79 MPa (242% strain) and high ionic conductivity of 0.32 S m-1 (0.07 M ZnSO4). Moreover, hydrogel sensor displayed high linear gauge factor 1.70 (0-200% strain), excellent stability, durability and reliability. The integrated hydrogel sensor also showed excellent sensor performance for human motion monitoring. This work provides a new prospect for the design of cellulose reinforced conductive hydrogels via a facile method.

Automated detection of severe diabetic retinopathy using deep learning method.

PURPOSE: The purpose of this study is to develop and validate the intelligent diagnosis of severe DR with lesion recognition based on color fundus photography. METHODS: The Kaggle public dataset for DR grading is used in the project, including 53,576 fundus photos in the test set, 28,101 in the training set, and 7,025 in the validation set. We randomly select 4,192 images for lesion annotation. Inception V3 structure is adopted as the classification algorithm. Both 299 × 299 pixel images and 896 × 896 pixel images are used as the input size. ROC curve, AUC, sensitivity, specificity, and their harmonic mean are used to evaluate the performance of the models. RESULTS: The harmonic mean and AUC of the model of 896 × 896 input are higher than those of the 299 × 299 input model. The sensitivity, specificity, harmonic mean, and AUC of the method with 896 × 896 resolution images as input for severe DR are 0.925, 0.907, 0.916, and 0.968, respectively. The prediction error mainly occurs in moderate NPDR, and cases with more hard exudates and cotton wool spots are easily predicted as severe cases. Cases with preretinal hemorrhage and vitreous hemorrhage are easily identified as severe cases, and IRMA is the most difficult lesion to recognize. CONCLUSIONS: We have studied the intelligent diagnosis of severe DR based on color fundus photography. This artificial intelligence-based technology offers a possibility to increase the accessibility and efficiency of severe DR screening.

Development and evaluation of a deep learning model for the detection of multiple fundus diseases based on colour fundus photography.

AIM: To explore and evaluate an appropriate deep learning system (DLS) for the detection of 12 major fundus diseases using colour fundus photography. METHODS: Diagnostic performance of a DLS was tested on the detection of normal fundus and 12 major fundus diseases including referable diabetic retinopathy, pathologic myopic retinal degeneration, retinal vein occlusion, retinitis pigmentosa, retinal detachment, wet and dry age-related macular degeneration, epiretinal membrane, macula hole, possible glaucomatous optic neuropathy, papilledema and optic nerve atrophy. The DLS was developed with 56 738 images and tested with 8176 images from one internal test set and two external test sets. The comparison with human doctors was also conducted. RESULTS: The area under the receiver operating characteristic curves of the DLS on the internal test set and the two external test sets were 0.950 (95% CI 0.942 to 0.957) to 0.996 (95% CI 0.994 to 0.998), 0.931 (95% CI 0.923 to 0.939) to 1.000 (95% CI 0.999 to 1.000) and 0.934 (95% CI 0.929 to 0.938) to 1.000 (95% CI 0.999 to 1.000), with sensitivities of 80.4% (95% CI 79.1% to 81.6%) to 97.3% (95% CI 96.7% to 97.8%), 64.6% (95% CI 63.0% to 66.1%) to 100% (95% CI 100% to 100%) and 68.0% (95% CI 67.1% to 68.9%) to 100% (95% CI 100% to 100%), respectively, and specificities of 89.7% (95% CI 88.8% to 90.7%) to 98.1% (95%CI 97.7% to 98.6%), 78.7% (95% CI 77.4% to 80.0%) to 99.6% (95% CI 99.4% to 99.8%) and 88.1% (95% CI 87.4% to 88.7%) to 98.7% (95% CI 98.5% to 99.0%), respectively. When compared with human doctors, the DLS obtained a higher diagnostic sensitivity but lower specificity. CONCLUSION: The proposed DLS is effective in diagnosing normal fundus and 12 major fundus diseases, and thus has much potential for fundus diseases screening in the real world.

Lens Opacities Classification System III-based artificial intelligence program for automatic cataract grading.

PURPOSE: To establish and validate an artificial intelligence (AI)-assisted automatic cataract grading program based on the Lens Opacities Classification System III (LOCS III). SETTING: Eye and Ear, Nose, and Throat Hospital, Fudan University, Shanghai, China. DESIGN: AI training. METHODS: Advanced deep-learning algorithms, including Faster R-CNN and ResNet, were applied to the localization and analysis of the region of interest. An internal dataset from the EENT Hospital of Fudan University and an external dataset from the Pujiang Eye Study were used for AI training, validation, and testing. The datasets were automatically labeled on the AI platform regarding the capture mode and cataract grading based on the LOCS III. RESULTS: The AI program showed reliable capture mode recognition, grading, and referral capability for nuclear and cortical cataract grading. In the internal and external datasets, 99.4% and 100% of automatic nuclear grading, respectively, had an absolute prediction error of ≤1.0, with a satisfactory referral capability (area under the curve [AUC]: 0.983 for the internal dataset; 0.977 for the external dataset); 75.0% (internal dataset) and 93.5% (external dataset) of the automatic cortical grades had an absolute prediction error of ≤1.0, with AUCs of 0.855 and 0.795 for referral, respectively. Good consistency was observed between automatic and manual grading when both nuclear and cortical cataracts were evaluated. However, automatic grading of posterior subcapsular cataracts was impractical. CONCLUSIONS: The AI program proposed in this study showed robust grading and diagnostic performance for both nuclear and cortical cataracts, based on LOCS III.

Valorization of Miscanthus × giganteus by γ-Valerolactone/H2O/FeCl3 system toward efficient conversion of cellulose and hemicelluloses.

γ-Valerolactone (GVL), a biomass-derived green chemical, offers an environmentally responsible solvent for conversion of lignocellulose to high value-added chemicals. Herein, we report a two-step process for directly producing cellulosic residual, furfural and lignin from Miscanthus × giganteus (M. × giganteus) bypassing the isolation of xylose, which exhibits promising advantage in energy reduction. The optimized pretreatment (100 mM FeCl3 at 160 °C for 60 min) induced significant xylan removal (98.4%), resulting in rugged fibre surface, thus leading to the peak cellulose conversion of 99.3%. Furfural yield in the second step reached to 76.6% after 100 mM FeCl3 catalyzed GVL/H2O treatment at 180 °C for 10 min without addition of any chemical. The extracted lignin showed representative structure (such as ß-O-4', ß-ß' linkages) and medium molecular weight (4275.5 g/mol). 79.6% of furfural can be recovered by distillation. This study proposes a systematic and energy efficient approach for maximizing biomass utilization.

An Optical Coherence Tomography-Based Deep Learning Algorithm for Visual Acuity Prediction of Highly Myopic Eyes After Cataract Surgery.

BACKGROUND: Due to complicated and variable fundus status of highly myopic eyes, their visual benefit from cataract surgery remains hard to be determined preoperatively. We therefore aimed to develop an optical coherence tomography (OCT)-based deep learning algorithms to predict the postoperative visual acuity of highly myopic eyes after cataract surgery. MATERIALS AND METHODS: The internal dataset consisted of 1,415 highly myopic eyes having cataract surgeries in our hospital. Another external dataset consisted of 161 highly myopic eyes from Heping Eye Hospital. Preoperative macular OCT images were set as the only feature. The best corrected visual acuity (BCVA) at 4 weeks after surgery was set as the ground truth. Five different deep learning algorithms, namely ResNet-18, ResNet-34, ResNet-50, ResNet-101, and Inception-v3, were used to develop the model aiming at predicting the postoperative BCVA, and an ensemble learning was further developed. The model was further evaluated in the internal and external test datasets. RESULTS: The ensemble learning showed the lowest mean absolute error (MAE) of 0.1566 logMAR and the lowest root mean square error (RMSE) of 0.2433 logMAR in the validation dataset. Promising outcomes in the internal and external test datasets were revealed with MAEs of 0.1524 and 0.1602 logMAR and RMSEs of 0.2612 and 0.2020 logMAR, respectively. Considerable sensitivity and precision were achieved in the BCVA < 0.30 logMAR group, with 90.32 and 75.34% in the internal test dataset and 81.75 and 89.60% in the external test dataset, respectively. The percentages of the prediction errors within ± 0.30 logMAR were 89.01% in the internal and 88.82% in the external test dataset. CONCLUSION: Promising prediction outcomes of postoperative BCVA were achieved by the novel OCT-trained deep learning model, which will be helpful for the surgical planning of highly myopic cataract patients.