Deep learning method can be used to automatically analyze electrocardiogram (ECG) data and rapidly implement arrhythmia classification, which provides significant clinical value for the early screening of arrhythmias. How to select arrhythmia features effectively under limited abnormal sample supervision is an urgent issue to address. This paper proposed an arrhythmia classification algorithm based on an adaptive multi-feature fusion network. The algorithm extracted RR interval features from ECG signals, employed one-dimensional convolutional neural network (1D-CNN) to extract time-domain deep features, employed Mel frequency cepstral coefficients (MFCC) and two-dimensional convolutional neural network (2D-CNN) to extract frequency-domain deep features. The features were fused using adaptive weighting strategy for arrhythmia classification. The paper used the arrhythmia database jointly developed by the Massachusetts Institute of Technology and Beth Israel Hospital (MIT-BIH) and evaluated the algorithm under the inter-patient paradigm. Experimental results demonstrated that the proposed algorithm achieved an average precision of 75.2%, an average recall of 70.1% and an average F ₁-score of 71.3%, demonstrating high classification accuracy and being able to provide algorithmic support for arrhythmia classification in wearable devices.
Humans ; Arrhythmias, Cardiac/diagnosis* ; Algorithms ; Electrocardiography/methods* ; Neural Networks, Computer ; Signal Processing, Computer-Assisted ; Deep Learning ; Classification Algorithms

Heart sounds are critical for early detection of cardiovascular diseases, yet existing studies mostly focus on traditional signal segmentation, feature extraction, and shallow classifiers, which often fail to sufficiently capture the dynamic and nonlinear characteristics of heart sounds, limit recognition of complex heart sound patterns, and are sensitive to data imbalance, resulting in poor classification performance. To address these limitations, this study proposes a novel heart sound classification method that integrates improved Mel-frequency cepstral coefficients (MFCC) for feature extraction with a convolutional neural network (CNN) and a deep Transformer model. In the preprocessing stage, a Butterworth filter is applied for denoising, and continuous heart sound signals are directly processed without segmenting the cardiac cycles, allowing the improved MFCC features to better capture dynamic characteristics. These features are then fed into a CNN for feature learning, followed by global average pooling (GAP) to reduce model complexity and mitigate overfitting. Lastly, a deep Transformer module is employed to further extract and fuse features, completing the heart sound classification. To handle data imbalance, the model uses focal loss as the objective function. Experiments on two public datasets demonstrate that the proposed method performs effectively in both binary and multi-class classification tasks. This approach enables efficient classification of continuous heart sound signals, provides a reference methodology for future heart sound research for disease classification, and supports the development of wearable devices and home monitoring systems.
Heart Sounds/physiology* ; Humans ; Algorithms ; Neural Networks, Computer ; Signal Processing, Computer-Assisted ; Deep Learning ; Cardiovascular Diseases/diagnosis* ; Classification Algorithms

OBJECTIVE: To train convolutional networks using multi-lead ECG data and classify new data accurately to provide reliable information for clinical diagnosis. METHODS: The data were pre-processed with a bandpass filter, and signal framing was adopted to adjust the data of different lengths to the same size to facilitate network training and prediction. The dataset was expanded by increasing the sample size to improve the detection rate of abnormal samples. A depth-wise separable convolution structure was used for more specific feature extraction for different channels of twelve-lead ECG data. We trained the two classifiers for each label using the improved DenseNet to classify different labels. RESULTS: The propose model showed an accuracy of 80.13% for distinguishing between normal and abnormal ECG with a sensitivity of 80.38%, a specificity of 79.91% and a F1 score of 79.35%. CONCLUSIONS The model proposed herein can rapidly and effectively classify the ECG data. The running time of a single dataset on GPU is 33.59 ms, which allows real-time prediction to meet the clinical requirements.
Algorithms ; Arrhythmias, Cardiac ; diagnosis ; Databases as Topic ; Electrocardiography ; classification ; methods ; Humans ; Neural Networks (Computer) ; Sensitivity and Specificity

OBJECTIVE: We propose a heartbeat-based end-to-end classification of arrhythmias to improve the classification performance for supraventricular ectopic beat (SVEB) and ventricular ectopic beat (VEB). METHODS: The ECG signals were preprocessed by heartbeat segmentation and heartbeat alignment. An arrhythmia classifier was constructed based on convolutional neural network, and the proposed loss function was used to train the classifier. RESULTS: The proposed algorithm was verified on MIT-BIH arrhythmia database. The AUC of the proposed loss function for SVEB and VEB reached 0.77 and 0.98, respectively. With the first 5 min segment as the local data, the diagnostic sensitivities for SVEB and VEB were 78.28% and 98.88%, respectively; when 0, 50, 100, and 150 samples were used as the local data, the diagnostic sensitivities for SVEB and VEB reached 82.25% and 93.23%, respectively. CONCLUSIONS The proposed method effectively reduces the negative impact of class-imbalance and improves the diagnostic sensitivities for SVEB and VEB, and thus provides a new solution for automatic arrhythmia classification.
Algorithms ; Arrhythmias, Cardiac ; classification ; diagnosis ; Electrocardiography ; Heart Rate ; Humans ; Neural Networks (Computer) ; Ventricular Premature Complexes ; classification ; diagnosis

The early diagnosis of children with autism spectrum disorders (ASD) is essential. Electroencephalography (EEG) is one of most commonly used neuroimaging techniques as the most accessible and informative method. In this study, approximate entropy (ApEn), sample entropy (SaEn), permutation entropy (PeEn) and wavelet entropy (WaEn) were extracted from EEGs of ASD child and a control group, and Student's -test was used to analyze between-group differences. Support vector machine (SVM) algorithm was utilized to build classification models for each entropy measure derived from different regions. Permutation test was applied in search for optimize subset of features, with which the SVM model achieved best performance. The results showed that the complexity of EEGs in children with autism was lower than that of the normal control group. Among all four entropies, WaEn got a better classification performance than others. Classification results vary in different regions, and the frontal lobe showed the best performance. After feature selection, six features were filtered out and the accuracy rate was increased to 84.55%, which can be convincing for assisting early diagnosis of autism.
Algorithms ; Autism Spectrum Disorder ; classification ; diagnosis ; Child ; Electroencephalography ; Entropy ; Humans ; Support Vector Machine

This article aims to identify four commonly applied herbs from Curcuma genus of Zingiberaceae family,namely Curcumae Radix( Yujin),Curcumae Rhizoma( Ezhu),Curcumae Longae Rhizoma( Jianghuang) and Wenyujin Rhizoma Concisum( Pianjianghuang). The odor fingerprints of those four herbal medicines were collected by electronic nose,respectively. Meanwhile,XGBoost algorithm was introduced to data analysis and discriminant model establishment,with four indexes for performance evaluation,including accuracy,precision,recall,and F-measure. The discriminant model was established by XGBoost with positive rate of returning to 166 samples in the training set and 69 samples in the test set were 99. 39% and 95. 65%,respectively. The top four of the contribution to the discriminant model were LY2/g CT,P40/1,LY2/Gh and LY2/LG,the least contributing sensor was T70/2. Compared with support vector machine,random forest and artificial neural network,XGBoost algorithms shows better identification capacity with higher recognition efficiency. The accuracy,precision,recall and F-measure of the XGBoost discriminant model forecast set were 95. 65%,95. 25%,93. 07%,93. 75%,respectively. The superiority of XGBoost in the identification of Curcuma herbs was verified. Obviously,this new method could not only be suitable for digitization and objectification of traditional Chinese medicine( TCM) odor indicators,but also achieve the identification of different TCM based on their odor fingerprint in electronic nose system. The introduction of XGBoost algorithm and more excellent algorithms provide more ideas for the application of electronic nose in data mining for TCM studies.
Algorithms ; Curcuma/classification* ; Discriminant Analysis ; Drugs, Chinese Herbal/analysis* ; Electronic Nose ; Medicine, Chinese Traditional ; Odorants/analysis* ; Plants, Medicinal/classification*

Neurocognitive deficits are frequently observed in patients with schizophrenia and major depressive disorder (MDD). The relations between cognitive features may be represented by neurocognitive graphs based on cognitive features, modeled as Gaussian Markov random fields. However, it is unclear whether it is possible to differentiate between phenotypic patterns associated with the differential diagnosis of schizophrenia and depression using this neurocognitive graph approach. In this study, we enrolled 215 first-episode patients with schizophrenia (FES), 125 with MDD, and 237 demographically-matched healthy controls (HCs). The cognitive performance of all participants was evaluated using a battery of neurocognitive tests. The graphical LASSO model was trained with a one-vs-one scenario to learn the conditional independent structure of neurocognitive features of each group. Participants in the holdout dataset were classified into different groups with the highest likelihood. A partial correlation matrix was transformed from the graphical model to further explore the neurocognitive graph for each group. The classification approach identified the diagnostic class for individuals with an average accuracy of 73.41% for FES vs HC, 67.07% for MDD vs HC, and 59.48% for FES vs MDD. Both of the neurocognitive graphs for FES and MDD had more connections and higher node centrality than those for HC. The neurocognitive graph for FES was less sparse and had more connections than that for MDD. Thus, neurocognitive graphs based on cognitive features are promising for describing endophenotypes that may discriminate schizophrenia from depression.
Adult ; Algorithms ; Depressive Disorder, Major ; classification ; diagnosis ; Endophenotypes ; analysis ; Female ; Humans ; Machine Learning ; Male ; Markov Chains ; Neuropsychological Tests ; Schizophrenia ; classification ; diagnosis ; Young Adult

To determine the reason for the different mortality trends of ischemic heart disease (IHD) for China between Global Burden of Disease (GBD) 2010 and GBD2013, and to improve garbage code (GC) redistribution. All data were obtained from the disease surveillance points system, and two proportions for assigning chronic pulmonary heart disease (PHD) as GC to IHD were from GBD2010 and GBD2013, which were different for years before 2004. By using the GBD2013 approach, the age-standard mortality rate (ASMR) increased by 100.21% in 1991, 44.81% in 1996, and 42.47% in 2000 in comparison with the GBD2010 approach. The different methods of chronic PHD redistribution impacted the trend of IHD mortality, which elevated it in the earlier 1990s by using the GBD2013 approach. Thus, improving the redistribution of GC as a key step in mortality statistics is important.
Algorithms ; China ; epidemiology ; Databases, Factual ; Global Burden of Disease ; statistics & numerical data ; Humans ; Models, Biological ; Myocardial Ischemia ; classification ; epidemiology ; mortality ; Population Surveillance ; Time Factors

BACKGROUND/AIMS: In assigning patients with chronic obstructive pulmonary disease (COPD) to subgroups according to the updated guidelines of the Global Initiative for Chronic Obstructive Lung Disease, discrepancies have been noted between the COPD assessment test (CAT) criteria and modified Medical Research Council (mMRC) criteria. We investigated the determinants of symptom and risk groups and sought to identify a better CAT criterion. METHODS: This retrospective study included COPD patients seen between June 20, 2012, and December 5, 2012. The CAT score that can accurately predict an mMRC grade > or = 2 versus < 2 was evaluated by comparing the area under the receiver operating curve (AUROC) and by classification and regression tree (CART) analysis. RESULTS: Among 428 COPD patients, the percentages of patients classif ied into subgroups A, B, C, and D were 24.5%, 47.2%, 4.2%, and 24.1% based on CAT criteria and 49.3%, 22.4%, 8.9%, and 19.4% based on mMRC criteria, respectively. More than 90% of the patients who met the mMRC criteria for the 'more symptoms group' also met the CAT criteria. AUROC and CART analyses suggested that a CAT score > or = 15 predicted an mMRC grade > or = 2 more accurately than the current CAT score criterion. During follow-up, patients with CAT scores of 10 to 14 did not have a different risk of exacerbation versus those with CAT scores < 10, but they did have a lower exacerbation risk compared to those with CAT scores of 15 to 19. CONCLUSIONS: A CAT score > or = 15 is a better indicator for the 'more symptoms group' in the management of COPD patients.
Aged ; Algorithms ; Area Under Curve ; *Decision Support Techniques ; Decision Trees ; Female ; Humans ; Lung/*physiopathology ; Male ; Middle Aged ; Predictive Value of Tests ; Pulmonary Disease, Chronic Obstructive/classification/*diagnosis/physiopathology ; ROC Curve ; Regression Analysis ; Reproducibility of Results ; Republic of Korea ; Retrospective Studies ; Risk Factors ; Severity of Illness Index

To solve the complex interaction problems of hepatitis disease classification, we proposed a lasso method (least absolute shrinkage and selection operator method) with feature interaction. First, lasso penalized function and hierarchical convex constraint were added to the interactive model which is newly defined. Then the model was solved with the convex optimal method combining Karush-Kuhn-Tucker (KKT) condition with generalized gradient descent. Finally, the sparse solution of the main effect features and interactive features were derived, and the classification model was implemented. The experiments were performed on two liver data sets and proved that features interaction contributed to the classification of liver diseases. The experimental results showed that the feature interaction lasso method was of strong explanatory ability, and its effectiveness and efficiency were superior to those of lasso, of all pair-wise lasso, support vector machine (SVM) method, K nearest neighbor (KNN) method, linear discriminant analysis (LDA) classification method, etc.
Algorithms ; Cluster Analysis ; Discriminant Analysis ; Humans ; Liver Diseases ; classification ; Support Vector Machine