Polysomnography (PSG) monitoring is an important method for clinical diagnosis of diseases such as insomnia, apnea and so on. In order to solve the problem of time-consuming and energy-consuming sleep stage staging of sleep disorder patients using manual frame-by-frame visual judgment PSG, this study proposed a deep learning algorithm model combining convolutional neural networks (CNN) and bidirectional gate recurrent neural networks (Bi GRU). A dynamic sparse self-attention mechanism was designed to solve the problem that gated recurrent neural networks (GRU) is difficult to obtain accurate vector representation of long-distance information. This study collected 143 overnight PSG data of patients from Shanghai Mental Health Center with sleep disorders, which were combined with 153 overnight PSG data of patients from the open-source dataset, and selected 9 electrophysiological channel signals including 6 electroencephalogram (EEG) signal channels, 2 electrooculogram (EOG) signal channels and a single mandibular electromyogram (EMG) signal channel. These data were used for model training, testing and evaluation. After cross validation, the accuracy was (84.0±2.0)%, and Cohen's kappa value was 0.77±0.50. It showed better performance than the Cohen's kappa value of physician score of 0.75±0.11. The experimental results show that the algorithm model in this paper has a high staging effect in different populations and is widely applicable. It is of great significance to assist clinicians in rapid and large-scale PSG sleep automatic staging.
ObjectiveThe aim of this study was to investigate the changes in peripheral blood metabolites and transcriptomes in patients with obstructive sleep apnea (OSA) and to assess their diagnostic value as biomarkers. MethodsIn this study, we utilized liquid chromatography-tandem mass spectrometry (LC-MS/MS) lipid-targeted metabolomics to compare the metabolic profiles of 30 OSA patients with those of 30 healthy controls, identifying differential lipid metabolites. Through Kyoto Encyclopedia of Genes and Genomes (KEGG) enrichment analysis, we determined that the glycerolipid metabolism pathway was significantly different. Furthermore, we conducted transcriptome analysis on peripheral blood mononuclear cells (PBMCs) from six OSA patients and six healthy controls to evaluate the expression of molecules related to the pathway. ResultsA total of 168 differential lipid metabolites were identified, with significant differences in the glycerolipid metabolism pathway between OSA patients and healthy controls. Transcriptome analysis revealed that glycerolipid metabolism-related molecules GPAT, AGPAT, and LPIN were under expressed in OSA patient PBMCs, suggesting that the glycerolipid metabolism pathway is suppressed in OSA patients. Additionally, diagnostic value analysis showed that GPAT and AGPAT had high AUC values, indicating their potential as biomarkers for OSA. ConclusionThe suppression of the glycerolipid metabolism pathway is closely related to the development of OSA, and the under expression of key genes in this pathway, such as GPAT, AGPAT, and LPIN, may be involved in the pathophysiological process of OSA. These findings not only provide a new perspective for understanding the pathogenesis of OSA but also offer new scientific evidence for the treatment of OSA from the perspective of glycerolipid metabolism regulation.
Aiming at the problem that the unbalanced distribution of data in sleep electroencephalogram(EEG) signals and poor comfort in the process of polysomnography information collection will reduce the model's classification ability, this paper proposed a sleep state recognition method using single-channel EEG signals (WKCNN-LSTM) based on one-dimensional width kernel convolutional neural networks(WKCNN) and long-short-term memory networks (LSTM). Firstly, the wavelet denoising and synthetic minority over-sampling technique-Tomek link (SMOTE-Tomek) algorithm were used to preprocess the original sleep EEG signals. Secondly, one-dimensional sleep EEG signals were used as the input of the model, and WKCNN was used to extract frequency-domain features and suppress high-frequency noise. Then, the LSTM layer was used to learn the time-domain features. Finally, normalized exponential function was used on the full connection layer to realize sleep state. The experimental results showed that the classification accuracy of the one-dimensional WKCNN-LSTM model was 91.80% in this paper, which was better than that of similar studies in recent years, and the model had good generalization ability. This study improved classification accuracy of single-channel sleep EEG signals that can be easily utilized in portable sleep monitoring devices.
Objective To analyze the causes of missed diagnosis of sleep apnea hypopnea syndrome ( SAHS) . Methods 42 missed diagnosed cases with SAHS from May 2009 to May 2011 were retrospectively analyzed and related literatures were reviewed. Results The SAHS patients often visited the doctors for complications of SAHS such as hypertension, diabetes mellitus, metabolic syndrome, etc. Clinical misdiagnosis rate was very high. Lack of specific symptoms during the day, complicated morbidities, and insufficient knowledge of SAHS led to the high misdiagnosis rate and the poor treatment effect of patients with SAHS. Conclusion Strengthening the educational propaganda of SAHS, detail medical history collection, and polysomnography monitoring ( PSG) as early as possible can help diagnose SAHS more accurately and reduce missed diagnosis.
ObjectiveTo provide the possibility to explain the relationship between genotype and phenotype, and to provide reference for the clinical treatment of Sleep-related hypermotor epilepsy (SHE). MethodsWe retrospectively analyzed the case data of the child (patient 1) diagnosed with SHE in the outpatient department of the Second Affiliated Hospital of Wenzhou Medical University in December 2017, and inquired about his family history and growth and development history. We learned that the father (patient 2) of the child had a history of epilepsy, and we also collected his medical history and growth and development history of patient 2. We carried out the basic physical examination for the two patients, and basic blood routine and blood biochemical indicators have also been done. In addition, electroencephalogram, Wechsler intelligence assessment and cranial magnetic resonance imaging were performed. After the diagnosis of patients 1 and 2, we treated them with antiepileptic drugs and make them long-term follow-up. What’more, we collected the peripheral blood of patient 1 and his father and mother, sequenced the gene, established phylogenetic tree for the mutation gene, and compared the homologous protein sequence to judge the conservation of the mutation. Moreover, in silico analysis was used to analyze the pathogenicity of the mutant gene. ResultsWe find a family with epilepsy, of whom patient 1 and his father are with epilepsy. Their clinical manifestations are atypical, and their seizures are all in sleep. After a long-term follow-up of two patients' drug treatments, it is found that patient 1 and patient 2 respond well to the drugs. Gene test shows that the mutations of DEPDC5 (c.484-1del c.484_485del) and KCNQ2 (c.1164A> T) are at the same site in both patient 1 and patient 2, and the mutation sites are first reported. What’more, the homologous protein alignment shows that the amino acids corresponding to the two mutant genes are highly conserved. ConclusionThis study mainly reports a family with sleep-related hypermotor epilepsy. Patients 1 and patient 2 have novel mutations of DEPDC5 and KCNQ2 genes. In the long-term follow-up of this study, it is found that the patients are effective the antiepileptic drugs.
Sleep stage classification is essential for clinical disease diagnosis and sleep quality assessment. Most of the existing methods for sleep stage classification are based on single-channel or single-modal signal, and extract features using a single-branch, deep convolutional network, which not only hinders the capture of the diversity features related to sleep and increase the computational cost, but also has a certain impact on the accuracy of sleep stage classification. To solve this problem, this paper proposes an end-to-end multi-modal physiological time-frequency feature extraction network (MTFF-Net) for accurate sleep stage classification. First, multi-modal physiological signal containing electroencephalogram (EEG), electrocardiogram (ECG), electrooculogram (EOG) and electromyogram (EMG) are converted into two-dimensional time-frequency images containing time-frequency features by using short time Fourier transform (STFT). Then, the time-frequency feature extraction network combining multi-scale EEG compact convolution network (Ms-EEGNet) and bidirectional gated recurrent units (Bi-GRU) network is used to obtain multi-scale spectral features related to sleep feature waveforms and time series features related to sleep stage transition. According to the American Academy of Sleep Medicine (AASM) EEG sleep stage classification criterion, the model achieved 84.3% accuracy in the five-classification task on the third subgroup of the Institute of Systems and Robotics of the University of Coimbra Sleep Dataset (ISRUC-S3), with 83.1% macro F1 score value and 79.8% Cohen’s Kappa coefficient. The experimental results show that the proposed model achieves higher classification accuracy and promotes the application of deep learning algorithms in assisting clinical decision-making.
Patients with autoimmune encephalitis are mainly characterized by behavioral, mental and motor abnormalities, neurological dysfunction, memory deficits and seizures. Different antibody types of autoimmune encephalitis its pathogenesis, clinical characteristics are different, in recent years found immune related epilepsy is closely related to autoimmune encephalitis, based on autoimmune encephalitis type is more, we choose more common autoimmune encephalitis, expounds its characteristics, to help clinical diagnosis.
ObjectiveTo investigate the renal impairment and the risk factors of renal impairment in patients with OSA. MethodsData from patients who underwent polysomnography (PSG) in our department from July 2022 to January 2023 were collected, totaling 178 cases. Based on the results of the polysomnography, the patients were divided into an OSA group (145 cases) and a non-OSA group (33 cases). According to the severity of the condition, the OSA group was further divided into mild OSA (21 cases), moderate OSA (28 cases), and severe OSA (96 cases). The Pearson correlation analysis was further conducted to analyze the relationships between serum urea nitrogen (BUN), serum cystatin C (Cys-C) concentrations, and estimated Glomerular Filtration Rate (eGFR) with various risk factors that may influence renal impairment. Moreover, multiple linear regression analysis was used to identify the risk factors affecting BUN, Cys-C, and eGFR. ResultsWhen comparing the two groups, there were statistically significant differences in age, weight, BMI, neck circumference, waist circumference, eGFR、Cys-C、BUN, LSaO2, CT90% (all P<0.05). Univariate analysis of variance was used to compare differences in BUN, Serum creatinine (SCr), Cys-C, and eGFR among patients with mild, moderate, and severe OSA, indicating that differences in eGFR and Cys-C among OSA patients of varying severities were statistically significant. Further analysis with Pearson correlation was conducted to explore the associations between eGFR, BUN, and Cys-C with potential risk factors that may affect renal function. Subsequently, multiple linear regression was utilized, taking these three indices as dependent variables to evaluate risk factors potentially influencing renal dysfunction. The results demonstrated that eGFR was negatively correlated with age, BMI, and CT90% (β=−0.95, P<0.001; β=−1.36, P=0.01; β=−32.64, P<0.001); BUN was positively correlated with CT90% (β=0.22, P=0.01); Cys-C was positively correlated with CT90% (β=0.58, P<0.001. Conclusion Chronic intermittent hypoxia, age, and obesity are risk factors for renal dysfunction in patients with OSA.
Objectives To study the characteristics and influencing factors of sleep disorder in patients with epilepsy. Methods One hundred and eighty-four patients with epilepsy who were admitted to the outpatient department and the epilepsy center in the Second Affiliated Hospital of Zhejiang University from October 2016 to October 2017 were enrolled. Their clinical data were collected in detail and their sleep related scales were evaluated. Sleep related assessment tools: Chinese version of the Pittsburgh sleep quality index scale (PSQI), the Epworth sleepiness scale (ESS), Berlin Questionnaire (BQ), Quality Of Life In People With Epilepsy-31 (QOLIE-31), Beck Anxiety Inventory (BAI) and Beck Depression Inventory(BDI). Results Among the 184 cases of patients with epilepsy, 100 cases were male (54.3%), 84 cases were female (45.7%), 35 cases (19.0%) had sleep disorders, 89 cases (48.4%) with poor quality of life, 23 cases (12.5%) with anxiety, 47 cases (25.5%) with depression, 59 cases (32.1%) had daytime sleepiness, and 30 cases (16.3%) with OSAS. there were statistically significant differences in age, history of hypertension, seizure frequency, quality of life , anxiety and depression in epilepsy patients with sleep disorder compared those without sleep disorder (P<0.05). The seizure frequency, quality of life, anxiety and depression were analyzed by logistic regression analysis, suggesting that seizure frequency (P=0.011) and depression (P<0.001) are independent risk factors of sleep disorders. Conclusions Epileptic patients with sleep disorder have higher frequency of seizures, poorer quality of life, and are more likely to be associated with anxiety and depression, and the frequency and depression are independent risk factors of sleep disorder in patients with epilepsy.
ObjectiveThe aim of this study was to investigate the value of Artificial Neural Networks (ANNs) in predicting the occurrence of Venous Thromboembolism (VTE) in patients with Obstructive Sleep Apnea (OSA), and to compare it with traditional Logistic regression models to assess its predictive efficacy, providing theoretical basis for the prediction of VTE risk in OSA patients. MethodsA retrospective analysis was conducted on patients diagnosed with OSA and hospitalized in the Department of Respiratory and Critical Care Medicine, Second Affiliated Hospital of Kunming Medical University, from January 2018 to August 2023. Patients were divided into OSA combined with VTE group (n=128) and pure OSA control group (n=680). The dataset was randomly divided into a training set (n=646) and an independent validation set (n=162). The Synthetic Minority Oversampling Technique (SMOTE) was employed to address the issue of data imbalance. Artificial Neural Networks and Logistic regression models were then built on training sets with and without SMOTE. Finally, the performance of each model was evaluated using accuracy, sensitivity, specificity, Youden's index, and Area Under the Receiver Operating Characteristic Curve (AUC). Results When oversampling was conducted using SMOTE on the training set, both the Artificial Neural Network and Logistic regression models showed improved AUC. The Artificial Neural Network model with SMOTE performed the best with an AUC value of 0.935 (95%CI: 0.898–0.961), achieving an accuracy of 90.15%, specificity of 87.32%, sensitivity of 93.44%, and Youden’s index of 0.808 at the optimal cutoff point. The Logistic regression model with SMOTE yielded an AUC value of 0.817 (95%CI: 0.765–0.861), with an accuracy of 77.27%, specificity of 83.80%, sensitivity of 69.67%, and Youden's index of 0.535. The difference in AUC between the Artificial Neural Network model and Logistic regression model was statistically significant after employing SMOTE (P<0.05). Conclusions The Artificial Neural Network model demonstrates high effectiveness in predicting VTE formation in OSA patients, particularly with the further improvement in predictive performance when utilizing SMOTE oversampling technique, rendering it more accurate and stable compared to the traditional Logistic regression model.