Search

Purpose: We aimed to develop an artificial intelligence-based myopic maculopathy grading method using EfficientNet to overcome the delayed grading and diagnosis of different myopic maculopathy degrees. Methods: The cooperative hospital provided 4642 healthy and myopic maculopathy color fundus photographs, comprising the four degrees of myopic maculopathy and healthy fundi. The myopic maculopathy grading models were trained using EfficientNet-B0 to EfficientNet-B7 models. The diagnostic results were compared with those of the VGG16 and ResNet50 classification models. The leading evaluation indicators were sensitivity, specificity, F1 score, area under the receiver operating characteristic (ROC) curve area under curve (AUC), 95% confidence interval, kappa value, and accuracy. The ROC curves of the ten grading models were also compared. Results: We used 1199 color fundus photographs to evaluate the myopic maculopathy grading models. The size of the EfficientNet-B0 myopic maculopathy grading model was 15.6 MB, and it had the highest kappa value (88.32%) and accuracy (83.58%). The model's sensitivities to diagnose tessellated fundus (TF), diffuse chorioretinal atrophy (DCA), patchy chorioretinal atrophy (PCA), and macular atrophy (MA) were 96.86%, 75.98%, 64.67%, and 88.75%, respectively. The specificity was above 93%, and the AUCs were 0.992, 0.960, 0.964, and 0.989, respectively. Conclusion: The EfficientNet models were used to design grading diagnostic models for myopic maculopathy. Based on the collected fundus images, the models could diagnose a healthy fundus and four types of myopic maculopathy. The models might help ophthalmologists to make preliminary diagnoses of different degrees of myopic maculopathy.

The accurate identification of tomato maturity and picking positions is essential for efficient picking. Current deep-learning models face challenges such as large parameter sizes, single-task limitations, and insufficient precision. This study proposes MTS-YOLO, a lightweight and efficient model for detecting tomato fruit bunch maturity and stem picking positions. We reconstruct the YOLOv8 neck network and propose the high- and low-level interactive screening path aggregation network (HLIS-PAN), which achieves excellent multi-scale feature extraction through the alternating screening and fusion of high- and low-level information while reducing the number of parameters. Furthermore, We utilize DySample for efficient upsampling, bypassing complex kernel computations with point sampling. Moreover, context anchor attention (CAA) is introduced to enhance the model’s ability to recognize elongated targets such as tomato fruit bunches and stems. Experimental results indicate that MTS-YOLO achieves an F1-score of 88.7% and an mAP@0.5 of 92.0%. Compared to mainstream models, MTS-YOLO not only enhances accuracy but also optimizes the model size, effectively reducing computational costs and inference time. The model precisely identifies the foreground targets that need to be harvested while ignoring background objects, contributing to improved picking efficiency. This study provides a lightweight and efficient technical solution for intelligent agricultural picking.

AimConventional approaches to diagnosing common eye diseases using B-mode ultrasonography are labor-intensive and time-consuming, must requiring expert intervention for accuracy. This study aims to address these challenges by proposing an intelligence-assisted analysis five-classification model for diagnosing common eye diseases using B-mode ultrasound images.MethodsThis research utilizes 2064 B-mode ultrasound images of the eye to train a novel model integrating artificial intelligence technology.ResultsThe ConvNeXt-L model achieved outstanding performance with an accuracy rate of 84.3% and a Kappa value of 80.3%. Across five classifications (no obvious abnormality, vitreous opacity, posterior vitreous detachment, retinal detachment, and choroidal detachment), the model demonstrated sensitivity values of 93.2%, 67.6%, 86.1%, 89.4%, and 81.4%, respectively, and specificity values ranging from 94.6% to 98.1%. F1 scores ranged from 71% to 92%, while AUC values ranged from 89.7% to 97.8%.ConclusionAmong various models compared, the ConvNeXt-L model exhibited superior performance. It effectively categorizes and visualizes pathological changes, providing essential assisted information for ophthalmologists and enhancing diagnostic accuracy and efficiency.

Purpose: Recently, the proportion of patients with high myopia has shown a continuous growing trend, more toward the younger age groups. This study aimed to predict the changes in spherical equivalent refraction (SER) and axial length (AL) in children using machine learning methods. Methods: This study is a retrospective study. The cooperative ophthalmology hospital of this study collected data on 179 sets of childhood myopia examinations. The data collected included AL and SER from grades 1 to 6. This study used the six machine learning models to predict AL and SER based on the data. Six evaluation indicators were used to evaluate the prediction results of the models. Results: For predicting SER in grade 6, grade 5, grade 4, grade 3, and grade 2, the best results were obtained through the multilayer perceptron (MLP) algorithm, MLP algorithm, orthogonal matching pursuit (OMP) algorithm, OMP algorithm, and OMP algorithm, respectively. The R2 of the five models were 0.8997, 0.7839, 0.7177, 0.5118, and 0.1758, respectively. For predicting AL in grade 6, grade 5, grade 4, grade 3, and grade 2, the best results were obtained through the Extra Tree (ET) algorithm, MLP algorithm, kernel ridge (KR) algorithm, KR algorithm, and MLP algorithm, respectively. The R2 of the five models were 0.7546, 0.5456, 0.8755, 0.9072, and 0.8534, respectively. Conclusion: Therefore, in predicting SER, the OMP model performed better than the other models in most experiments. In predicting AL, the KR and MLP models were better than the other models in most experiments.

Abstract Background PSMA-TO-1 (“Tumor-Optimized-1”) is a novel PSMA ligand with longer circulation time than PSMA-617. We compared the biodistribution in subcutaneous tumor-bearing mice of PSMA-TO-1, PSMA-617 and PSMA-11 when labeled with 68Ga and 177Lu, and the survival after treatment with 225Ac-PSMA-TO-1/-617 in a murine model of disseminated prostate cancer. We also report dosimetry data of 177Lu-PSMA-TO1/-617 in prostate cancer patients. Methods First, PET images of 68Ga-PSMA-TO-1/-617/-11 were acquired on consecutive days in three mice bearing subcutaneous C4-2 xenografts. Second, 50 subcutaneous tumor-bearing mice received either 30 MBq of 177Lu-PSMA-617 or 177Lu-PSMA-TO-1 and were sacrificed at 1, 4, 24, 48 and 168 h for ex vivo gamma counting and biodistribution. Third, mice bearing disseminated lesions via intracardiac inoculation were treated with either 40 kBq of 225Ac-PSMA-617, 225Ac-PSMA-TO-1, or remained untreated and followed for survival. Additionally, 3 metastatic castration-resistant prostate cancer patients received 500 MBq of 177Lu-PSMA-TO-1 under compassionate use for dosimetry purposes. Planar images with an additional SPECT/CT acquisition were acquired for dosimetry calculations. Results Tumor uptake measured by PET imaging of 68Ga-labeled agents in mice was highest using PSMA-617, followed by PSMA-TO-1 and PSMA-11. 177Lu-PSMA tumor uptake measured by ex vivo gamma counting at subsequent time points tended to be greater for PSMA-TO-1 up to 1 week following treatment (p > 0.13 at all time points). This was, however, accompanied by increased kidney uptake and a 26-fold higher kidney dose of PSMA-TO-1 compared with PSMA-617 in mice. Mice treated with a single-cycle 225Ac-PSMA-TO-1 survived longer than those treated with 225Ac-PSMA-617 and untreated mice, respectively (17.8, 14.5 and 7.7 weeks, respectively; p

Abstract A simple method was proposed to activate alkaline Cu(OH)2 with an acidic ionomer, Nafion, to regulate its surface microenvironment, including hydrophobicity and local basicity. In particular, the direct complete neutralization reaction between Cu(OH)2 and Nafion in aqueous solution induces the exposing of vast anions which can exclude the in‐situ‐formed hydroxides and raise the local basicity. Remarkably, the optimal Nafion‐activated Cu(OH)2‐derived Cu can efficiently suppress the hydrogen evolution reaction (HER) and improve the selectivity for multi‐carbon products in the CO2 electroreduction reaction (eCO2RR). The H2 Faradaic efficiency (FE) decreased to 11% at a current density of 300 mA/cm2 (−0.76 V vs. RHE) in a flow cell, while the bare one with H2 had an FE of 40%. The total eCO2RR FE reaches as high as 83%, along with an evidently increased C2H4 FE of 44% as compared with the bare one (24%), and good stability (8000 s), surpassing that of most of the reported Cu(OH)2‐derived Cu. The experimental and theoretical results both show that the strong hydrophobicity and high local basicity jointly boosted the eCO2RR as acquired by felicitously introducing ionomer on the Cu(OH)2‐derived Cu surface.

In the field of neuroscience, it is very important to evaluate the causal coupling characteristics between bioelectrical signals accurately and effectively. Transfer entropy is commonly used to analyze complex data, especially the causal relationship between data with non-linear, multidimensional characteristics. However, traditional transfer entropy needs to estimate the probability density function of the variable, which is computationally complex and unstable. In this paper, a new and effective method for entropy transfer is proposed, by means of applying R-vine copula function estimation. The effectiveness of R-vine copula transfer entropy is first verified on several simulations, and then applied to intermuscular coupling analysis to explore the characteristics of the intermuscular coupling network of muscles in non-fatigue and fatigue conditions. The experiment results show that as the muscle group enters the fatigue state, the community structure can be adjusted and the muscle nodes participating in the exercise are fully activated, enabling the two-way interaction between different communities. Finally, it comes to the conclusion that the proposed method can make accurate inferences about complex causal coupling. Moreover, the characteristics of the intermuscular coupling network in both non-fatigue and fatigue states can provide a new theoretical perspective for the diagnosis of neuromuscular fatigue and sports rehabilitation, which has good application value.

PurposeTo assess the value of an automated classification model for dry and wet macular degeneration based on the ConvNeXT model.MethodsA total of 672 fundus images of normal, dry, and wet macular degeneration were collected from the Affiliated Eye Hospital of Nanjing Medical University and the fundus images of dry macular degeneration were expanded. The ConvNeXT three-category model was trained on the original and expanded datasets, and compared to the results of the VGG16, ResNet18, ResNet50, EfficientNetB7, and RegNet three-category models. A total of 289 fundus images were used to test the models, and the classification results of the models on different datasets were compared. The main evaluation indicators were sensitivity, specificity, F1-score, area under the curve (AUC), accuracy, and kappa.ResultsUsing 289 fundus images, three-category models trained on the original and expanded datasets were assessed. The ConvNeXT model trained on the expanded dataset was the most effective, with a diagnostic accuracy of 96.89%, kappa value of 94.99%, and high diagnostic consistency. The sensitivity, specificity, F1-score, and AUC values for normal fundus images were 100.00, 99.41, 99.59, and 99.80%, respectively. The sensitivity, specificity, F1-score, and AUC values for dry macular degeneration diagnosis were 87.50, 98.76, 90.32, and 97.10%, respectively. The sensitivity, specificity, F1-score, and AUC values for wet macular degeneration diagnosis were 97.52, 97.02, 96.72, and 99.10%, respectively.ConclusionThe ConvNeXT-based category model for dry and wet macular degeneration automatically identified dry and wet macular degeneration, aiding rapid, and accurate clinical diagnosis.

PurposeTo assess the value of automatic disc-fovea angle (DFA) measurement using the DeepLabv3+ segmentation model.MethodsA total of 682 normal fundus image datasets were collected from the Eye Hospital of Nanjing Medical University. The following parts of the images were labeled and subsequently reviewed by ophthalmologists: optic disc center, macular center, optic disc area, and virtual macular area. A total of 477 normal fundus images were used to train DeepLabv3+, U-Net, and PSPNet model, which were used to obtain the optic disc area and virtual macular area. Then, the coordinates of the optic disc center and macular center were obstained by using the minimum outer circle technique. Finally the DFA was calculated.ResultsIn this study, 205 normal fundus images were used to test the model. The experimental results showed that the errors in automatic DFA measurement using DeepLabv3+, U-Net, and PSPNet segmentation models were 0.76°, 1.4°, and 2.12°, respectively. The mean intersection over union (MIoU), mean pixel accuracy (MPA), average error in the center of the optic disc, and average error in the center of the virtual macula obstained by using DeepLabv3+ model was 94.77%, 97.32%, 10.94 pixels, and 13.44 pixels, respectively. The automatic DFA measurement using DeepLabv3+ got the less error than the errors that using the other segmentation models. Therefore, the DeepLabv3+ segmentation model was finally chosen to measure DFA automatically.ConclusionsThe DeepLabv3+ segmentation model -based automatic segmentation techniques can produce accurate and rapid DFA measurements.

Background: Artificial intelligence (AI) has been used in the research of ophthalmic disease diagnosis, and it may have an impact on medical and ophthalmic practice in the future. This study explores the general application and research frontier of artificial intelligence in ophthalmic disease detection.Methods: Citation data were downloaded from the Web of Science Core Collection database to evaluate the extent of the application of Artificial intelligence in ophthalmic disease diagnosis in publications from 1 January 2012, to 31 December 2021. This information was analyzed using CiteSpace.5.8. R3 and Vosviewer.Results: A total of 1,498 publications from 95 areas were examined, of which the United States was determined to be the most influential country in this research field. The largest cluster labeled “Brownian motion” was used prior to the application of AI for ophthalmic diagnosis from 2007 to 2017, and was an active topic during this period. The burst keywords in the period from 2020 to 2021 were system, disease, and model.Conclusion: The focus of artificial intelligence research in ophthalmic disease diagnosis has transitioned from the development of AI algorithms and the analysis of abnormal eye physiological structure to the investigation of more mature ophthalmic disease diagnosis systems. However, there is a need for further studies in ophthalmology and computer engineering.

Abstract Background Definitive radiation therapy (dRT) is an effective initial treatment of intermediate-risk (IR) and high-risk (HR) prostate cancer (PCa). PSMA PET/CT is superior to standard of care imaging (CT, MRI, bone scan) for detecting regional and distant metastatic PCa. PSMA PET/CT thus has the potential to guide patient selection and the planning for dRT and improve patient outcomes. Methods This is a multicenter randomized phase 3 trial (NCT04457245). We will randomize 312 patients to proceed with standard dRT (control Arm, n = 150), or undergo a PSMA PET/CT scan at the study site (both 18F-DCFPyL and 68Ga-PSMA-11 can be used) prior to dRT planning (intervention arm, n = 162). dRT will be performed at the treating radiation oncologist facility. In the control arm, dRT will be performed as routinely planned. In the intervention arm, the treating radiation oncologist can incorporate PSMA PET/CT findings into the RT planning. Androgen deprivation therapy (ADT) is administered per discretion of the treating radiation oncologist and may be modified as a result of the PSMA PET/CT results. We assume that approximately 8% of subjects randomized to the PSMA PET arm will be found to have M1 disease and thus will be more appropriate candidates for long-term systemic or multimodal therapy, rather than curative intent dRT. PET M1 patients will thus not be included in the primary endpoint analysis. The primary endpoint is the success rate of patients with unfavorable IR and HR PCa after standard dRT versus PSMA PET-based dRT. Secondary Endpoints (whole cohort) include progression free survival (PFS), metastasis-free survival after initiation of RT, overall survival (OS), % of change in initial treatment intent and Safety. Discussion This is the first randomized phase 3 prospective trial designed to determine whether PSMA PET/CT molecular imaging can improve outcomes in patients with PCa who receive dRT. In this trial the incorporation of PSMA PET/CT may improve the success rate of curative intent radiotherapy in two ways: to optimize patient selection as a biomarker and to personalizes the radiotherapy plan. Clinical trial registration UCLA IND#147591 ○ Submission: 02.27.2020 ○ Safe-to-proceed letter issued by FDA: 04.01.2020 UCLA IRB #20–000378 ClinicalTrials.gov Identifier NCT04457245 . Date of Registry: 07.07.2020. Essen EudraCT 2020–003526-23

Abstract Endoscopic mucosal resection (EMR) and endoscopic submucosal dissection (ESD) are well-established therapeutics for gastrointestinal neoplasias, but complications after EMR/ESD, including bleeding and perforation, result in additional treatment morbidity and even threaten the lives of patients. Thus, designing biomaterials to treat gastric bleeding and wound healing after endoscopic treatment is highly desired and remains a challenge. Herein, a series of injectable pH-responsive self-healing adhesive hydrogels based on acryloyl-6-aminocaproic acid (AA) and AA-g-N-hydroxysuccinimide (AA-NHS) were developed, and their great potential as endoscopic sprayable bioadhesive materials to efficiently stop hemorrhage and promote the wound healing process was further demonstrated in a swine gastric hemorrhage/wound model. The hydrogels showed a suitable gelation time, an autonomous and efficient self-healing capacity, hemostatic properties, and good biocompatibility. With the introduction of AA-NHS as a micro-cross-linker, the hydrogels exhibited enhanced adhesive strength. A swine gastric hemorrhage in vivo model demonstrated that the hydrogels showed good hemostatic performance by stopping acute arterial bleeding and preventing delayed bleeding. A gastric wound model indicated that the hydrogels showed excellent treatment effects with significantly enhanced wound healing with type I collagen deposition, α-SMA expression, and blood vessel formation. These injectable self-healing adhesive hydrogels exhibited great potential to treat gastric wounds after endoscopic treatment.

Abstract Background Left ventricular mural thrombus (LVMT) is a life-threatening complication in patients with left ventricular dysfunction. Case presentation A 67-year-old man had a history of penetrating myocardial infarction and left ventricular aneurysm (LVA). The patient was scheduled for a non-cardiac surgery and stopped aspirin for 10 days to reduce the risk of bleeding. Fresh LVMT was revealed via the transesophageal echocardiography (TEE) after the preoperative discontinuation of aspirin. Conclusions Perioperative repeated evaluation for the thrombosis by echocardiography is essential in cases of patients with cardiovascular disease undergoing non-cardiac surgery. In high risk patient, during temporary interruption of antiplatelets, bridging with perioperative low-molecular-weight heparin is advisable.

PurposeA six-category model of common retinal diseases is proposed to help primary medical institutions in the preliminary screening of the five common retinal diseases.MethodsA total of 2,400 fundus images of normal and five common retinal diseases were provided by a cooperative hospital. Two six-category deep learning models of common retinal diseases based on the EfficientNet-B4 and ResNet50 models were trained. The results from the six-category models in this study and the results from a five-category model in our previous study based on ResNet50 were compared. A total of 1,315 fundus images were used to test the models, the clinical diagnosis results and the diagnosis results of the two six-category models were compared. The main evaluation indicators were sensitivity, specificity, F1-score, area under the curve (AUC), 95% confidence interval, kappa and accuracy, and the receiver operator characteristic curves of the two six-category models were compared in the study.ResultsThe diagnostic accuracy rate of EfficientNet-B4 model was 95.59%, the kappa value was 94.61%, and there was high diagnostic consistency. The AUC of the normal diagnosis and the five retinal diseases were all above 0.95. The sensitivity, specificity, and F1-score for the diagnosis of normal fundus images were 100, 99.9, and 99.83%, respectively. The specificity and F1-score for RVO diagnosis were 95.68, 98.61, and 93.09%, respectively. The sensitivity, specificity, and F1-score for high myopia diagnosis were 96.1, 99.6, and 97.37%, respectively. The sensitivity, specificity, and F1-score for glaucoma diagnosis were 97.62, 99.07, and 94.62%, respectively. The sensitivity, specificity, and F1-score for DR diagnosis were 90.76, 99.16, and 93.3%, respectively. The sensitivity, specificity, and F1-score for MD diagnosis were 92.27, 98.5, and 91.51%, respectively.ConclusionThe EfficientNet-B4 model was used to design a six-category model of common retinal diseases. It can be used to diagnose the normal fundus and five common retinal diseases based on fundus images. It can help primary doctors in the screening for common retinal diseases, and give suitable suggestions and recommendations. Timely referral can improve the efficiency of diagnosis of eye diseases in rural areas and avoid delaying treatment.

Retinal images are the most intuitive medical images for the diagnosis of fundus diseases. Low-quality retinal images cause difficulties in computer-aided diagnosis systems and the clinical diagnosis of ophthalmologists. The high quality of retinal images is an important basis of precision medicine in ophthalmology. In this study, we propose a retinal image enhancement method based on deep learning to enhance multiple low-quality retinal images. A generative adversarial network is employed to build a symmetrical network, and a convolutional block attention module is introduced to improve the feature extraction capability. The retinal images in our dataset are sorted into two sets according to their quality: low and high quality. Generators and discriminators alternately learn the features of low/high-quality retinal images without the need for paired images. We analyze the proposed method both qualitatively and quantitatively on public datasets and a private dataset. The study results demonstrate that the proposed method is superior to other advanced algorithms, especially in enhancing color-distorted retinal images. It also performs well in the task of retinal vessel segmentation. The proposed network effectively enhances low-quality retinal images, aiding ophthalmologists and enabling computer-aided diagnosis in pathological analysis. Our method enhances multiple types of low-quality retinal images using a deep learning network.

In order to reduce the computational tasks in robots with large-scale and complex knowledge, several methods of robotic knowledge localization have been proposed over the past decades. Logic is an important and useful tool for complex robotic reasoning, action planning, learning and verification. This paper uses propositional atoms in logic to describe the affecting factors of robotic large-scale knowledge. Definability in logic reasoning shows that truths of some propositional atoms are decided by other propositional atoms. Definability technology is an important method to eliminate inessential propositional atoms in robotic large-scale and complex knowledge, so the computational tasks in robotic knowledge can be completed faster. On the other hand, by applying the splitting technique, the knowledge base can be equivalently divided into a number of sub-knowledge bases, without sharing any propositional atoms with others. In this paper, we show that the inessential propositional atoms can be decided faster by the local definability technology based on the splitting method, first formed in local belief revision by Parikh in 1999. Hence, the decision-making in robotic large-scale and complex knowledge is more effective.

CD147 is correlated with tumor aggressiveness in various human malignancies. Here, we investigated CD147 protein expression in 223 patients with gastric cancer by immunohistochemistry and analyzed its association with disease-free and overall survival. CD147 was increased in gastric cancer compared to normal tissues. Additionally, CD147 expression was associated with gastric cancer invasion, metastasis and TNM stage, whereas it was not related to age, sex, differentiation status, tumor site or Lauren classification. Kaplan-Meier analysis confirmed that CD147 was associated with disease-free and overall survival in patients with gastric cancer; i.e., patients with positive CD147 staining tend to have worse disease-free and overall survival. Moreover, Cox's proportional hazards analysis demonstrated that CD147 was an independent marker of disease-free and overall survival for patients with gastric cancer. These results confirm the association of CD147 with gastric cancer invasion and metastasis and prove that CD147 might be an indicator of tumor recurrence and prognosis in gastric cancer.

There is much discussion in the cancer drug development community about how to incorporate molecular tools into early-stage clinical trials to assess target modulation, measure anti-tumor activity, and enrich the clinical trial population for patients who are more likely to benefit. Small, molecularly focused clinical studies offer the promise of the early definition of optimal biologic dose and patient population.Based on preclinical evidence that phosphatase and tensin homolog deleted on Chromosome 10 (PTEN) loss sensitizes tumors to the inhibition of mammalian target of rapamycin (mTOR), we conducted a proof-of-concept Phase I neoadjuvant trial of rapamycin in patients with recurrent glioblastoma, whose tumors lacked expression of the tumor suppressor PTEN. We aimed to assess the safety profile of daily rapamycin in patients with glioma, define the dose of rapamycin required for mTOR inhibition in tumor tissue, and evaluate the antiproliferative activity of rapamycin in PTEN-deficient glioblastoma. Although intratumoral rapamycin concentrations that were sufficient to inhibit mTOR in vitro were achieved in all patients, the magnitude of mTOR inhibition in tumor cells (measured by reduced ribosomal S6 protein phosphorylation) varied substantially. Tumor cell proliferation (measured by Ki-67 staining) was dramatically reduced in seven of 14 patients after 1 wk of rapamycin treatment and was associated with the magnitude of mTOR inhibition (p = 0.0047, Fisher exact test) but not the intratumoral rapamycin concentration. Tumor cells harvested from the Ki-67 nonresponders retained sensitivity to rapamycin ex vivo, indicating that clinical resistance to biochemical mTOR inhibition was not cell-intrinsic. Rapamycin treatment led to Akt activation in seven patients, presumably due to loss of negative feedback, and this activation was associated with shorter time-to-progression during post-surgical maintenance rapamycin therapy (p < 0.05, Logrank test).Rapamycin has anticancer activity in PTEN-deficient glioblastoma and warrants further clinical study alone or in combination with PI3K pathway inhibitors. The short-term treatment endpoints used in this neoadjuvant trial design identified the importance of monitoring target inhibition and negative feedback to guide future clinical development.http://www.ClinicalTrials.gov (#NCT00047073).