Application of artificial intelligence in ophthalmic plastic surgery / 国际眼科杂志(Guoji Yanke Zazhi)

The advancement of computers and data explosion have ushered in the third wave of artificial intelligence(AI). AI is an interdisciplinary field that encompasses new ideas, new theories, and new technologies, etc. AI has brought convenience to ophthalmology application and promoted its intelligent, precise, and minimally invasive development. At present, AI has been widely applied in various fields of ophthalmology, especially in oculoplastic surgery. AI has made rapid progress in image detection, facial recognition, etc., and its performance and accuracy have even surpassed humans in some aspects. This article reviews the relevant research and applications of AI in oculoplastic surgery, including ptosis, single eyelid, pouch, eyelid mass, and exophthalmos, and discusses the challenges and opportunities faced by AI in oculoplastic surgery, and provides prospects for its future development, aiming to provide new ideas for the development of AI in oculoplastic surgery.

Transfer learning enhanced graph neural network for aldehyde oxidase metabolism prediction and its experimental application

Aldehyde oxidase (AOX) is a molybdoenzyme that is primarily expressed in the liver and is involved in the metabolism of drugs and other xenobiotics. AOX-mediated metabolism can result in unexpected outcomes, such as the production of toxic metabolites and high metabolic clearance, which can lead to the clinical failure of novel therapeutic agents. Computational models can assist medicinal chemists in rapidly evaluating the AOX metabolic risk of compounds during the early phases of drug discovery and provide valuable clues for manipulating AOX-mediated metabolism liability. In this study, we developed a novel graph neural network called AOMP for predicting AOX-mediated metabolism. AOMP integrated the tasks of metabolic substrate/non-substrate classification and metabolic site prediction, while utilizing transfer learning from 13C nuclear magnetic resonance data to enhance its performance on both tasks. AOMP significantly outperformed the benchmark methods in both cross-validation and external testing. Using AOMP, we systematically assessed the AOX-mediated metabolism of common fragments in kinase inhibitors and successfully identified four new scaffolds with AOX metabolism liability, which were validated through in vitro experiments. Furthermore, for the convenience of the community, we established the first online service for AOX metabolism prediction based on AOMP, which is freely available at https://aomp.alphama.com.cn.

Optimization of ethanol extraction process for Yihuang powder by orthogonal experiment combined with AHP and GA-BP neural network / 中国药房

OBJECTIVE To optimize ethanol extraction process of Yihuang powder. METHODS An orthogonal experiment was designed by reflux extraction with ethanol volume fraction， liquid-to-material ratio， and extraction time as investigation factors. The parameters used were the contents of hesperidin， nobiletin， tangeretin， gallic acid， chebulagic acid， chebulinic acid， liquiritin， glycyrrhizin， eugenol， and the paste-forming rate. The analytic hierarchy process （AHP） was used to calculate the comprehensive score. The optimal ethanol extraction process parameters of Yihuang powder were determined by verifying the results predicted by orthogonal experiment and genetic algorithm （GA）-back propagation neural network （BP neural network）. RESULTS The optimal ethanol extraction process parameters， as optimized by orthogonal experiment， were as follows: ethanol volume fraction of 60%， liquid-solid ratio of 14∶1 （mL/g)， extraction time of 90 min， and extraction for 2 times. The comprehensive score obtained by verification was 79.19. Meanwhile， the optimal ethanol extraction process parameters， optimized by GA-BP neural network， were ethanol volume fraction of 65%， liquid-solid ratio of 14∶1 （mL/g )， extraction time of 60 min， and extraction for 2 times. The comprehensive score obtained by verification was 85.30， higher than the results obtained from orthogonal experiment. CONCLUSIONS The optimization method of orthogonal experiment combined with GA-BP neural network is superior to the traditional orthogonal experiment optimization method. The optimized ethanol extraction process of Yihuang powder is stable and reliable.

Optimization of water extraction technology of Xiangqin jiere granules by orthogonal design based on G1-entropy weight compared with BP neural network / 中国药房

OBJECTIVE Optimizing the water extraction technology of Xiangqin jiere granules. METHODS The orthogonal test of 3 factors and 3 levels was designed， and comprehensive scoring was conducted for the above indexes by using G1-entropy weight to obtain the optimized water extraction technology of Xiangqin jiere granules with water addition ratio， extraction time and extraction times as factors， using the contents of forsythoside A， baicalin， phillyrin， oroxylin A-7-O-β-D-glycoside， wogonoside， baicalein and wogonin， and extraction rate as evaluation indexes. BP neural network modeling was used to optimize the network model and water extraction process using the results of 9 groups of orthogonal tests as test and training data， the water addition multiple， decocting time and extraction times as input nodes， and the comprehensive score as output nodes. Then the two analysis methods were compared by verification test to find the best water extraction process parameters. RESULTS The water extraction technology optimized by the orthogonal test was 8-fold water， extracting 3 times， extracting for 1 h each time. Comprehensive score was 96.84 （RSD＝0.90%）. The optimal water extraction technology obtained by BP neural network modeling included 12-fold water， extracting 4 times， extracting for 0.5 h each time. The comprehensive score was 92.72 （RSD＝0.77%）， which was slightly lower than that of the orthogonal test. CONCLUSIONS The water extraction technology of Xiangqin jiere granules is optimized successfully in the study， which includes adding 8-fold water， extracting 3 times， and extracting for 1 hour each time.

The Effectiveness of Machine Learning Systems' Accuracy in Predicting Heart Stroke Using Socio-Demographic and Risk Factors - A Comparative Analysis of Various Models

Background: Cardiologists can more appropriately classify patients' cardiovascular diseases by executing ac-curate diagnoses and prognoses, enabling them to administer the most appropriate care. Due to machine learning's ability to identify patterns in data, its applications in the medical sector have grown. Diagnosticians can avoid making mistakes by classifying the incidence of cardiovascular illness using machine learning. To lower the fatality rate brought on by cardiovascular disorders, our research developed a model that can cor-rectly forecast these conditions.Methods: This study emphasized a model that can correctly forecast cardiovascular illnesses to lower the death rate brought on by these conditions. We deployed four well-known classification machine learning al-gorithms like K nearest Neighbour, Logistic Regression, Artificial Neural network, and Decision tree. Results: The proposed models were evaluated by their performance matrices. However logistic regression performed high accuracy concerning AUC (0.955) 95% CI (0.872-0.965) followed by the artificial neural net-work. AUC (0.864) 95% CI (0.826-0.912). Conclusion: Individuals' risk of having a cardiac event may be predicted using machine learning, and those who are most at risk can be identified. Predictive models may be developed via machine learning to pinpoint those who have a high chance of suffering a heart attack

Construction of hiPSC-derived Excitatory Neural Network-like Tissue / 中山大学学报(医学科学版)

ObjectiveTo construct a neural network-like tissue with the potential of synaptic formation in vitro by seeding human induced pluripotent stem cell-derived neural precursor cells （hiPSC-NPCs） on decellularized optic nerve （DON）， so as to provide a promising approach for repair of nerve tissue injury. MethodsThrough directional induction and tissue engineering technology， human induced pluripotent stem cells （hiPSCs） and 3D DON scaffolds were combined to construct neural network-like tissues. Then the hiPSCs were directionally induced into human neural precursor cells （hNPCs） and neurons. Immunofluorescence staining was used to identify cell differentiation efficiency. 3D DON scaffolds were prepared. Morphology and cytocompatibility of scaffolds were identified by scanning electron microscopy and Tunnel staining. Induced hiPSC-NPCs were seeded on DON scaffolds. Immunofluorescence staining， scanning electron microscopy， transmission electron microscopy and patch clamp were used to observe the morphology and functional identification of constructed neural network tissues. Results①The results of immunofluorescence staining suggested that most of hiPSC-NPCs differentiated into neurons in vitro. We had successfully constructed a neural network dominated by neurons. ② The results of scanning electron microscopy and immunohistochemistry suggested that a neural network-like tissue with predominating excitatory neurons in vitro was successfully constructed. ③The results of immunohistochemical staining， transmission electron microscopy and patch clamp indicated that the neural network-like tissue had synaptic transmission function. ConclusionA neural network-like tissue mainly composed of excitatory neurons has been constructed by the combination of natural uniform-channel DON scaffold and hiPSC-NPCs， which has the function of synaptic transmission. This neural network plays a significant role in stem cell derived replacement therapy， and offers a promising prospect for repair of spinal cord injury （SCI） and other neural tissue injuries.

Screening of natural drug molecules against Fusarium oxysporum of ginseng root rot based on machine learning / 药学学报

italic>Fusarium oxysporum widely exists in farmland soil and is one of the main pathogenic fungi of root rot, which seriously affects the growth and development of plants and often causes serious losses of cash crops. In order to screen out natural compounds that inhibit the activity of Fusarium oxysporum more economically and efficiently, random forest, support vector machine and artificial neural network based on machine learning algorithms were constructed using the information of known inhibitory compounds in ChEMBL database in this study. And the antibacterial activity of the screened drugs was verified thereafter. The results showed that the prediction accuracy of the three models reached 77.58%, 83.03% and 81.21%, respectively. Based on the inhibition experiment, the best inhibition effect (MIC = 0.312 5 mg·mL-1) of ononin was verified. The virtual screening method proposed in this study provides ideas for the development and creation of new pesticides derived from natural products, and the screened ononin is expected to be a potential lead compound for the development of novel inhibitors of Fusarium oxysporum.

Risk predictive models of healthcare-seeking delay among imported malaria patients in Jiangsu Province based on the machine learning / 中国血吸虫病防治杂志

Objective To create risk predictive models of healthcare-seeking delay among imported malaria patients in Jiangsu Province based on machine learning algorithms, so as to provide insights into early identification of imported malaria cases in Jiangsu Province. Methods Case investigation, first symptoms and time of initial diagnosis of imported malaria patients in Jiangsu Province in 2019 were captured from Infectious Disease Report Information Management System and Parasitic Disease Prevention and Control Information Management System of Chinese Center for Disease Control and Prevention. The risk predictive models of healthcare-seeking delay among imported malaria patients were created with the back propagation (BP) neural network model, logistic regression model, random forest model and Bayesian model using thirteen factors as independent variables, including occupation, species of malaria parasite, main clinical manifestations, presence of complications, severity of disease, age, duration of residing abroad, frequency of malaria parasite infections abroad, incubation period, level of institution at initial diagnosis, country of origin, number of individuals travelling with patients and way to go abroad, and time of healthcare-seeking delay as a dependent variable. Logistic regression model was visualized using a nomogram, and the nomogram was evaluated using calibration curves. In addition, the efficiency of the four models for prediction of risk of healthcare-seeking delay among imported malaria patients was evaluated using the area under curve (AUC) of receiver operating characteristic curve (ROC). The importance of each characteristic was quantified and attributed by using SHAP to examine the positive and negative effects of the value of each characteristic on the predictive efficiency. Results A total of 244 imported malaria patients were enrolled, including 100 cases (40.98%) with the duration from onset of first symptoms to time of initial diagnosis that exceeded 24 hours. Logistic regression analysis identified a history of malaria parasite infection [odds ratio (OR) = 3.075, 95% confidential interval (CI): (1.597, 5.923)], long incubation period [OR = 1.010, 95% CI: (1.001, 1.018)] and seeking healthcare in provincial or municipal medical facilities [OR = 12.550, 95% CI: (1.158, 135.963)] as risk factors for delay in seeking healthcare among imported malaria cases. BP neural network modeling showed that duration of residing abroad, incubation period and age posed great impacts on delay in healthcare-seek among imported malaria patients. Random forest modeling showed that the top five factors with the greatest impact on healthcare-seeking delay included main clinical manifestations, the way to go abroad, incubation period, duration of residing abroad and age among imported malaria patients, and Bayesian modeling revealed that the top five factors affecting healthcare-seeking delay among imported malaria patients included level of institutions at initial diagnosis, age, country of origin, history of malaria parasite infection and individuals travelling with imported malaria patients. ROC curve analysis showed higher overall performance of the BP neural network model and the logistic regression model for prediction of the risk of healthcare-seeking delay among imported malaria patients (Z = 2.700 to 4.641, all P values < 0.01), with no statistically significant difference in the AUC among four models (Z = 1.209, P > 0.05). The sensitivity (71.00%) and Youden index (43.92%) of the logistic regression model was higher than those of the BP neural network (63.00% and 36.61%, respectively), and the specificity of the BP neural network model (73.61%) was higher than that of the logistic regression model (72.92%). Conclusions Imported malaria cases with long duration of residing abroad, a history of malaria parasite infection, long incubation period, advanced age and seeking healthcare in provincial or municipal medical institutions have a high likelihood of delay in healthcare-seeking in Jiangsu Province. The models created based on the logistic regression and BP neural network show a high efficiency for prediction of the risk of healthcare-seeking among imported malaria patients in Jiangsu Province, which may provide insights into health management of imported malaria patients.

Intelligent identification of livestock, a source of Schistosoma japonicum infection, based on deep learning of unmanned aerial vehicle images / 中国血吸虫病防治杂志

Objective To develop an intelligent recognition model based on deep learning algorithms of unmanned aerial vehicle (UAV) images, and to preliminarily explore the value of this model for remote identification, monitoring and management of cattle, a source of Schistosoma japonicum infection. Methods Oncomelania hupensis snail-infested marshlands around the Poyang Lake area were selected as the study area. Image datasets of the study area were captured by aerial photography with UAV and subjected to augmentation. Cattle in the sample database were annotated with the annotation software VGG Image Annotator to create the morphological recognition labels for cattle. A model was created for intelligent recognition of livestock based on deep learning-based Mask R-convolutional neural network (CNN) algorithms. The performance of the model for cattle recognition was evaluated with accuracy, precision, recall, F1 score and mean precision. Results A total of 200 original UAV images were obtained, and 410 images were yielded following data augmentation. A total of 2 860 training samples of cattle recognition were labeled. The created deep learning-based Mask R-CNN model converged following 200 iterations, with an accuracy of 88.01%, precision of 92.33%, recall of 94.06%, F1 score of 93.19%, and mean precision of 92.27%, and the model was effective to detect and segment the morphological features of cattle. Conclusion The deep learning-based Mask R-CNN model is highly accurate for recognition of cattle based on UAV images, which is feasible for remote intelligent recognition, monitoring, and management of the source of S. japonicum infection.

Enzymatic Digestion Method Coupled with Artificial Intelligence Techniques in Forensic Drowning Diatom Detection / 中山大学学报(医学科学版)

ObjectiveArtificial intelligence （AI） full smear automated diatom detection technology can perform forensic pathology drowning diatom detection more quickly and efficiently than human experts.However， this technique was only used in conjunction with the strong acid digestion method， which has a low extraction rate of diatoms. In this study， we propose to use the more efficient proteinase K tissue digestion method （hereinafter referred to as enzyme digestion method） as a diatom extraction method to investigate the generalization ability and feasibility of this technique in other diatom extraction methods. MethodsLung tissues from 6 drowned cadavers were collected for proteinase K ablation and made into smears， and the smears were digitized using the digital image matrix cutting method and a diatom and background database was established accordingly.The data set was divided into training set， validation set and test set in the ratio of 3：1：1， and the convolutional neural network （CNN） models were trained， internally validated， and externally tested on the basis of ImageNet pre-training. ResultsThe results showed that the accuracy rate of the external test of the best model was 97.65 %， and the area where the model features were extracted was the area where the diatoms were located. The best CNN model in practice had a precision of more than 80 % for diatom detection of drowned corpses. ConclusionIt is shown that the AI automated diatom detection technique based on CNN model and enzymatic digestion method in combination can efficiently identify diatoms and can be used as an auxiliary method for diatom detection in drowning identification.

Treatment Strategies and Prognostic Factors in Patients with Stage T3 and T4 Laryngeal Carcinoma / 肿瘤防治研究

Objective To investigate the selection of treatment strategies and prognostic factors for patients with stage T3 and T4 laryngeal carcinoma. Methods A total of 132 patients with stage T3 and T4 laryngeal cancer admitted to our hospital from March 2010 to March 2019 were retrospectively selected as research objects. According to the different treatment strategies, the patients were divided into simple surgery group (group A, 57 cases), simple chemoradiotherapy group (group B, 32 cases), and surgery combined with chemoradiotherapy group (group C, 43 cases). The general data and clinicopathological features of the three groups were compared, and a survival curve was drawn by the Kaplan–Meier method. The 3-year survival rates of the three groups were compared. Then, the same 132 patients were divided into survival and death groups. The clinical data of the two groups were compared, and the prognostic factors were analyzed by multivariate logistic regression. A back propagation (BP) neural network model was constructed, and its differentiation and accuracy were evaluated. Results The proportions and 3 year survival rates of patients with poor differentiation, lymphatic vascular invasion, and involvement of lymph nodes outside the capsule in group C were significantly higher than those in groups A and B (P < 0.05). The 3 year survival rate of 132 patients was 68.94%(41/132). Poor differentiation, N2-N3 stage, lymphatic vascular invasion, and involvement of lymph nodes outside the capsule were risk factors for death (P < 0.05), whereas surgery combined with radiotherapy and chemotherapy were protective factors (P < 0.05). The BP neural network model exhibited good discrimination and high accuracy. Conclusion Surgery combined with radiotherapy and chemotherapy can significantly improve survival rate in patients with poor differentiation, lymphatic vascular invasion, and involvement of lymph nodes outside the capsule. Close attention should be paid to patients with stage N2-N3 in the formulation of reasonable treatment strategies.

Application of Deep Learning in Differential Diagnosis of Ameloblastoma and Odontogenic Keratocyst Based on Panoramic Radiographs / 中国医学科学院学报

Objective To evaluate the accuracy of different convolutional neural networks (CNN),representative deep learning models,in the differential diagnosis of ameloblastoma and odontogenic keratocyst,and subsequently compare the diagnosis results between models and oral radiologists. Methods A total of 1000 digital panoramic radiographs were retrospectively collected from the patients with ameloblastoma (500 radiographs) or odontogenic keratocyst (500 radiographs) in the Department of Oral and Maxillofacial Radiology,Peking University School of Stomatology.Eight CNN including ResNet (18,50,101),VGG (16,19),and EfficientNet (b1,b3,b5) were selected to distinguish ameloblastoma from odontogenic keratocyst.Transfer learning was employed to train 800 panoramic radiographs in the training set through 5-fold cross validation,and 200 panoramic radiographs in the test set were used for differential diagnosis.Chi square test was performed for comparing the performance among different CNN.Furthermore,7 oral radiologists (including 2 seniors and 5 juniors) made a diagnosis on the 200 panoramic radiographs in the test set,and the diagnosis results were compared between CNN and oral radiologists. Results The eight neural network models showed the diagnostic accuracy ranging from 82.50% to 87.50%,of which EfficientNet b1 had the highest accuracy of 87.50%.There was no significant difference in the diagnostic accuracy among the CNN models (P=0.998,P=0.905).The average diagnostic accuracy of oral radiologists was (70.30±5.48)%,and there was no statistical difference in the accuracy between senior and junior oral radiologists (P=0.883).The diagnostic accuracy of CNN models was higher than that of oral radiologists (P<0.001). Conclusion Deep learning CNN can realize accurate differential diagnosis between ameloblastoma and odontogenic keratocyst with panoramic radiographs,with higher diagnostic accuracy than oral radiologists.

Prediction of pulp exposure risk of carious pulpitis based on deep learning / 华西口腔医学杂志

OBJECTIVES@#This study aims to predict the risk of deep caries exposure in radiographic images based on the convolutional neural network model, compare the prediction results of the network model with those of senior dentists, evaluate the performance of the model for teaching and training stomatological students and young dentists, and assist dentists to clarify treatment plans and conduct good doctor-patient communication before surgery.@*METHODS@#A total of 206 cases of pulpitis caused by deep caries were selected from the Department of Stomatological Hospital of Tianjin Medical University from 2019 to 2022. According to the inclusion and exclusion criteria, 104 cases of pulpitis were exposed during the decaying preparation period and 102 cases of pulpitis were not exposed. The 206 radiographic images collected were randomly divided into three groups according to the proportion: 126 radiographic images in the training set, 40 radiographic images in the validation set, and 40 radiographic images in the test set. Three convolutional neural networks, visual geometry group network (VGG), residual network (ResNet), and dense convolutional network (DenseNet) were selected to analyze the rules of the radiographic images in the training set. The radiographic images of the validation set were used to adjust the super parameters of the network. Finally, 40 radiographic images of the test set were used to evaluate the performance of the three network models. A senior dentist specializing in dental pulp was selected to predict whether the deep caries of 40 radiographic images in the test set were exposed. The gold standard is whether the pulp is exposed after decaying the prepared hole during the clinical operation. The prediction effect of the three network models (VGG, ResNet, and DenseNet) and the senior dentist on the pulp exposure of 40 radiographic images in the test set were compared using receiver operating characteristic (ROC) curve, area under the ROC curve (AUC), accuracy, sensitivity, specificity, positive predictive value, negative predictive value, and F1 score to select the best network model.@*RESULTS@#The best network model was DenseNet model, with AUC of 0.97. The AUC values of the ResNet model, VGG model, and the senior dentist were 0.89, 0.78, and 0.87, respectively. Accuracy was not statistically different between the senior dentist (0.850) and the DenseNet model (0.850)(P>0.05). Kappa consistency test showed moderate reliability (Kappa=0.6>0.4, P<0.05).@*CONCLUSIONS@#Among the three convolutional neural network models, the DenseNet model has the best predictive effect on whether deep caries are exposed in imaging. The predictive effect of this model is equivalent to the level of senior dentists specializing in dental pulp.

Heart sound classification algorithm based on time-frequency combination feature and adaptive fuzzy neural network / 生物医学工程学杂志

Feature extraction methods and classifier selection are two critical steps in heart sound classification. To capture the pathological features of heart sound signals, this paper introduces a feature extraction method that combines mel-frequency cepstral coefficients (MFCC) and power spectral density (PSD). Unlike conventional classifiers, the adaptive neuro-fuzzy inference system (ANFIS) was chosen as the classifier for this study. In terms of experimental design, we compared different PSDs across various time intervals and frequency ranges, selecting the characteristics with the most effective classification outcomes. We compared four statistical properties, including mean PSD, standard deviation PSD, variance PSD, and median PSD. Through experimental comparisons, we found that combining the features of median PSD and MFCC with heart sound systolic period of 100-300 Hz yielded the best results. The accuracy, precision, sensitivity, specificity, and F1 score were determined to be 96.50%, 99.27%, 93.35%, 99.60%, and 96.35%, respectively. These results demonstrate the algorithm's significant potential for aiding in the diagnosis of congenital heart disease.

The inverse stochastic resonance in a small-world neuronal network under electromagnetic stimulation / 生物医学工程学杂志

Electromagnetic stimulation is an important neuromodulation technique that modulates the electrical activity of neurons and affects cortical excitability for the purpose of modulating the nervous system. The phenomenon of inverse stochastic resonance is a response mechanism of the biological nervous system to external signals and plays an important role in the signal processing of the nervous system. In this paper, a small-world neural network with electrical synaptic connections was constructed, and the inverse stochastic resonance of the small-world neural network under electromagnetic stimulation was investigated by analyzing the dynamics of the neural network. The results showed that: the Levy channel noise under electromagnetic stimulation could cause the occurrence of inverse stochastic resonance in small-world neural networks; the characteristic index and location parameter of the noise had significant effects on the intensity and duration of the inverse stochastic resonance in neural networks; the larger the probability of randomly adding edges and the number of nearest neighbor nodes in small-world networks, the more favorable the anti-stochastic resonance was; by adjusting the electromagnetic stimulation parameters, a dual regulation of the inverse stochastic resonance of the neural network can be achieved. The results of this study provide some theoretical support for exploring the regulation mechanism of electromagnetic nerve stimulation technology and the signal processing mechanism of nervous system.

Alzheimer's disease classification based on nonlinear high-order features and hypergraph convolutional neural network / 生物医学工程学杂志

Alzheimer's disease (AD) is an irreversible neurodegenerative disorder that damages patients' memory and cognitive abilities. Therefore, the diagnosis of AD holds significant importance. The interactions between regions of interest (ROIs) in the brain often involve multiple areas collaborating in a nonlinear manner. Leveraging these nonlinear higher-order interaction features to their fullest potential contributes to enhancing the accuracy of AD diagnosis. To address this, a framework combining nonlinear higher-order feature extraction and three-dimensional (3D) hypergraph neural networks is proposed for computer-assisted diagnosis of AD. First, a support vector machine regression model based on the radial basis function kernel was trained on ROI data to obtain a base estimator. Then, a recursive feature elimination algorithm based on the base estimator was applied to extract nonlinear higher-order features from functional magnetic resonance imaging (fMRI) data. These features were subsequently constructed into a hypergraph, leveraging the complex interactions captured in the data. Finally, a four-dimensional (4D) spatiotemporal hypergraph convolutional neural network model was constructed based on the fMRI data for classification. Experimental results on the Alzheimer's Disease Neuroimaging Initiative (ADNI) database demonstrated that the proposed framework outperformed the Hyper Graph Convolutional Network (HyperGCN) framework by 8% and traditional two-dimensional (2D) linear feature extraction methods by 12% in the AD/normal control (NC) classification task. In conclusion, this framework demonstrates an improvement in AD classification compared to mainstream deep learning methods, providing valuable evidence for computer-assisted diagnosis of AD.

A bio-inspired hierarchical spiking neural network with biological synaptic plasticity for event camera object recognition / 生物医学工程学杂志

With inherent sparse spike-based coding and asynchronous event-driven computation, spiking neural network (SNN) is naturally suitable for processing event stream data of event cameras. In order to improve the feature extraction and classification performance of bio-inspired hierarchical SNNs, in this paper an event camera object recognition system based on biological synaptic plasticity is proposed. In our system input event streams were firstly segmented adaptively using spiking neuron potential to improve computational efficiency of the system. Multi-layer feature learning and classification are implemented by our bio-inspired hierarchical SNN with synaptic plasticity. After Gabor filter-based event-driven convolution layer which extracted primary visual features of event streams, we used a feature learning layer with unsupervised spiking timing dependent plasticity (STDP) rule to help the network extract frequent salient features, and a feature learning layer with reward-modulated STDP rule to help the network learn diagnostic features. The classification accuracies of the network proposed in this paper on the four benchmark event stream datasets were better than the existing bio-inspired hierarchical SNNs. Moreover, our method showed good classification ability for short event stream input data, and was robust to input event stream noise. The results show that our method can improve the feature extraction and classification performance of this kind of SNNs for event camera object recognition.

MF-SuP-pKa: Multi-fidelity modeling with subgraph pooling mechanism for pKa prediction

Acid-base dissociation constant (pKa) is a key physicochemical parameter in chemical science, especially in organic synthesis and drug discovery. Current methodologies for pKa prediction still suffer from limited applicability domain and lack of chemical insight. Here we present MF-SuP-pKa (multi-fidelity modeling with subgraph pooling for pKa prediction), a novel pKa prediction model that utilizes subgraph pooling, multi-fidelity learning and data augmentation. In our model, a knowledge-aware subgraph pooling strategy was designed to capture the local and global environments around the ionization sites for micro-pKa prediction. To overcome the scarcity of accurate pKa data, low-fidelity data (computational pKa) was used to fit the high-fidelity data (experimental pKa) through transfer learning. The final MF-SuP-pKa model was constructed by pre-training on the augmented ChEMBL data set and fine-tuning on the DataWarrior data set. Extensive evaluation on the DataWarrior data set and three benchmark data sets shows that MF-SuP-pKa achieves superior performances to the state-of-the-art pKa prediction models while requires much less high-fidelity training data. Compared with Attentive FP, MF-SuP-pKa achieves 23.83% and 20.12% improvement in terms of mean absolute error (MAE) on the acidic and basic sets, respectively.

Progress in neural network mechanism of tinnitus using functional magnetic resonance imaging / 临床耳鼻咽喉头颈外科杂志

Tinnitus refers to the perception of abnormal sound in the absence of external sound stimulation. It can have an impact on a person's mood, memory, attention, and mental state, although the mechanism of tinnitus is still unclear. In recent years, the research on the central neural mechanism of tinnitus has attracted the attention of scholars.Functional magnetic resonance imaging （fMRI）,as an effective imaging technology, has been actively employed in this field. This paper provides a systematic summary of studies on the central neural mechanism of tinnitus by fMRI in recent years,revealed the changes of functional connections among tinnitus-related neural networks,such as auditory network,limbic system,default mode network and salience network. The central neural mechanism of tinnitus involves multiple networks that interact with each other. By understanding this mechanism, we hope to develop more targeted prevention and treatment strategies to help patients alleviate long-term tinnitus.

Diagnosis of nasopharyngeal carcinoma with convolutional neural network on narrowband imaging / 临床耳鼻咽喉头颈外科杂志

Objective:To evaluate the diagnostic accuracy of the convolutional neural network（CNN） in diagnosing nasopharyngeal carcinoma using endoscopic narrowband imaging. Methods:A total of 834 cases with nasopharyngeal lesions were collected from the People's Hospital of Guangxi Zhuang Autonomous Region between 2014 and 2016. We trained the DenseNet201 model to classify the endoscopic images, evaluated its performance using the test dataset, and compared the results with those of two independent endoscopic experts. Results:The area under the ROC curve of the CNN in diagnosing nasopharyngeal carcinoma was 0.98. The sensitivity and specificity of the CNN were 91.90% and 94.69%, respectively. The sensitivity of the two expert-based assessment was 92.08% and 91.06%, respectively, and the specificity was 95.58% and 92.79%, respectively. There was no significant difference between the diagnostic accuracy of CNN and the expert-based assessment （P=0.282, P=0.085）. Moreover, there was no significant difference in the accuracy in discriminating early-stage and late-stage nasopharyngeal carcinoma（P=0.382）. The CNN model could rapidly distinguish nasopharyngeal carcinoma from benign lesions, with an image recognition time of 0.1 s/piece. Conclusion:The CNN model can quickly distinguish nasopharyngeal carcinoma from benign nasopharyngeal lesions, which can aid endoscopists in diagnosing nasopharyngeal lesions and reduce the rate of nasopharyngeal biopsy.