A method to detect sleep apnea using residual attention mechanism network from single-lead ECG signal.

Sleep apnea is a sleep disorder caused by weakened or suspended breathing during sleep, which seriously affects the work and health of patients. The traditional polysomnography (PSG) detection process is complicated and expensive, which has attracted researchers to explore a rapid detection method based on single-lead ECG signals. However, existing ECG-based sleep apnea detection methods have certain limitations and complexities, mainly relying on human-crafted features. To solve the problem, the paper develops a sleep apnea detection method based on a residual attention mechanism network. The method uses the RR interval signal and the R-peak signal derived from the ECG signal as input, realizes feature extraction through the residual network (ResNet), and adds the SENet attention mechanism to deepen the mining of channel features. Experimental results show that the per-segment accuracy of the proposed method can reach 86.2%. Compared with existing works, its accuracy has increased by 1.1-8.1%. These results show that the proposed residual attention network can effectively use ECG signals to quickly detect sleep apnea. Meanwhile, compared with existing works, the proposed method overcomes the limitations and complexity of human-crafted features in sleep apnea detection research.

SMRT sequencing of full-length transcriptome and gene expression analysis in two chemical types of Pogostemon cablin (Blanco) Benth.

BACKGROUND: Pogostemon cablin (Blanco) Benth. also called patchouli, is a traditional medicinal and aromatic plant that grows mainly in Southeast Asia and China. In China, P. cablin is divided into two chemical types: the patchouliol-type and the pogostone-type. Patchouliol-type patchouli usually grow taller, with thicker stems and bigger leaves, and produce more aromatic oil. METHODS: To better understand the genetic differences between the two chemical types that contribute to their differences in morphology and biosynthetic capabilities, we constructed de novo transcriptomes from both chemical types using the Pacific Biosciences (PacBio) Sequel platform and performed differential expression analysis of multiple tissues using Illumina short reads. RESULTS: In this study, using single-molecule real-time (SMRT) long-read sequencing, we obtained 22.07 GB of clean data and 134,647 nonredundant transcripts from two chemical types. Additionally, we identified 126,576 open reading frames (ORFs), 100,638 coding sequences (CDSs), 4,106 long noncoding RNAs (lncRNAs) and 6,829 transcription factors (TFs) from two chemical types of P. cablin. We adopted PacBio and Illumina sequencing to identify differentially expressed transcripts (DEGs) in three tissues of the two chemical types. More DEGs were observed in comparisons of different tissues collected from the same chemical type relative to comparisons of the same tissue collected from different chemical types. Furthormore, using KEGG enrichment analysis of DEGs, we found that the most enriched biosynthetic pathways of secondary metabolites of the two chemical types were "terpenoid backbone biosynthesis", "phenylpropanoid biosynthesis", "plant hormone signal transduction", "sesquiterpenoid and triterpenoid biosynthesis", "ubiquinone and other terpenoid-quinone biosynthesis", "flavonoid biosynthesis", and "flavone and flavonol biosynthesis". However, the main pathways of the patchouliol-type also included "diterpene biosynthesis" and "monoterpene biosynthesis". Additionally, by comparing the expression levels of the three tissues verified by qRT-PCR, more DEGs in the roots were upregulated in the mevalonate (MVA) pathway in the cytoplasm, but more DEGs in the leaves were upregulated in the methylerythritol phosphate (MEP) pathway in the plastid, both of which are important pathways for terpenoids biosynthesis. These findings promote the study of further genome annotation and transcriptome research in P. cablin.

Sleep Apnea Detection Based on Multi-Scale Residual Network.

Aiming at the fact that traditional convolutional neural networks cannot effectively extract signal features in complex application scenarios, a sleep apnea (SA) detection method based on multi-scale residual networks is proposed. First, we analyze the physiological mechanism of SA, which uses the RR interval signals and R peak signals derived from the ECG signals as input. Then, a multi-scale residual network is used to extract the characteristics of the original signals in order to obtain sensitive characteristics from various angles. Because the residual structure is used in the model, the problem of model degradation can be avoided. Finally, a fully connected layer is introduced for SA detection. In order to overcome the impact of class imbalance, a focal loss function is introduced to replace the traditional cross-entropy loss function, which makes the model pay more attention to learning difficult samples in the training phase. Experimental results from the Apnea-ECG dataset show that the accuracy, sensitivity and specificity of the proposed multi-scale residual network are 86.0%, 84.1% and 87.1%, respectively. These results indicate that the proposed method not only achieves greater recognition accuracy than other methods, but it also effectively resolves the problem of low sensitivity caused by class imbalance.

Automatic ECG Classification Using Continuous Wavelet Transform and Convolutional Neural Network.

Early detection of arrhythmia and effective treatment can prevent deaths caused by cardiovascular disease (CVD). In clinical practice, the diagnosis is made by checking the electrocardiogram (ECG) beat-by-beat, but this is usually time-consuming and laborious. In the paper, we propose an automatic ECG classification method based on Continuous Wavelet Transform (CWT) and Convolutional Neural Network (CNN). CWT is used to decompose ECG signals to obtain different time-frequency components, and CNN is used to extract features from the 2D-scalogram composed of the above time-frequency components. Considering the surrounding R peak interval (also called RR interval) is also useful for the diagnosis of arrhythmia, four RR interval features are extracted and combined with the CNN features to input into a fully connected layer for ECG classification. By testing in the MIT-BIH arrhythmia database, our method achieves an overall performance of 70.75%, 67.47%, 68.76%, and 98.74% for positive predictive value, sensitivity, F1-score, and accuracy, respectively. Compared with existing methods, the overall F1-score of our method is increased by 4.75~16.85%. Because our method is simple and highly accurate, it can potentially be used as a clinical auxiliary diagnostic tool.

PGNet: Pipeline Guidance for Human Key-Point Detection.

Human key-point detection is a challenging research field in computer vision. Convolutional neural models limit the number of parameters and mine the local structure, and have made great progress in significant target detection and key-point detection. However, the features extracted by shallow layers mainly contain a lack of semantic information, while the features extracted by deep layers contain rich semantic information but a lack of spatial information that results in information imbalance and feature extraction imbalance. With the complexity of the network structure and the increasing amount of computation, the balance between the time of communication and the time of calculation highlights the importance. Based on the improvement of hardware equipment, network operation time is greatly improved by optimizing the network structure and data operation methods. However, as the network structure becomes deeper and deeper, the communication consumption between networks also increases, and network computing capacity is optimized. In addition, communication overhead is also the focus of recent attention. We propose a novel network structure PGNet, which contains three parts: pipeline guidance strategy (PGS); Cross-Distance-IoU Loss (CIoU); and Cascaded Fusion Feature Model (CFFM).

Intercropping With Turmeric or Ginger Reduce the Continuous Cropping Obstacles That Affect Pogostemon cablin (Patchouli).

Continuous cropping (CC) restricts the development of the medicinal plant cultivation industry because it alters soil properties and the soil microbial micro-ecological environment. It can also lead to reductions in the chemical contents of medicinal plants. In this study, we intercropped continuously cropped Pogostemon cablin (patchouli) with turmeric or ginger. High-throughput sequencing was used to study the soil bacteria and fungi. Community composition, diversity, colony structure, and colony differences were also analyzed. A redundancy analysis (RDA) was used to study the interactions between soil physical and chemical factors, and the bacteria and fungi. The correlations between the soil community and the soil physical and chemical properties were also investigated. The results showed that intercropping turmeric and ginger with patchouli can improve soil microbial abundance, diversity, and community structure by boosting the number of dominant bacteria, and by improving soil bacterial metabolism and the activities of soil enzymes. They also modify the soil physical and chemical properties through changes in enzyme activity, soil pH, and soil exchangeable Ca (Ca). In summary, turmeric and ginger affect the distribution of dominant bacteria, and increase the contents of the active ingredient in patchouli. The results from this study suggested that the problems associated with continuously cropping patchouli can be ameliorated by intercropping it with turmeric and ginger.

A hybrid method for heartbeat classification via convolutional neural networks, multilayer perceptrons and focal loss.

BACKGROUND: Heart arrhythmia, as one of the most important cardiovascular diseases (CVDs), has gained wide attention in the past two decades. The article proposes a hybrid method for heartbeat classification via convolutional neural networks, multilayer perceptrons and focal loss. METHODS: In the method, a convolution neural network is used to extract the morphological features. The reason behind this is that the morphological characteristics of patients have inter-patient variations, which makes it difficult to accurately describe using traditional hand-craft ways. Then the extracted morphological features are combined with the RR intervals features and input into the multilayer perceptron for heartbeat classification. The RR intervals features contain the dynamic information of the heartbeat. Furthermore, considering that the heartbeat classes are imbalanced and would lead to the poor performance of minority classes, a focal loss is introduced to resolve the problem in the article. RESULTS: Tested using the MIT-BIH arrhythmia database, our method achieves an overall positive predictive value of 64.68%, sensitivity of 68.55%, f1-score of 66.09%, and accuracy of 96.27%. Compared with existing works, our method significantly improves the performance of heartbeat classification. CONCLUSIONS: Our method is simple yet effective, which is potentially used for personal automatic heartbeat classification in remote medical monitoring. The source code is provided on https://github.com/JackAndCole/Deep-Neural-Network-For-Heartbeat-Classification.

Sleep apnea detection from a single-lead ECG signal with automatic feature-extraction through a modified LeNet-5 convolutional neural network.

Sleep apnea (SA) is the most common respiratory sleep disorder, leading to some serious neurological and cardiovascular diseases if left untreated. The diagnosis of SA is traditionally made using Polysomnography (PSG). However, this method requires many electrodes and wires, as well as an expert to monitor the test. Several researchers have proposed instead using a single channel signal for SA diagnosis. Among these options, the ECG signal is one of the most physiologically relevant signals of SA occurrence, and one that can be easily recorded using a wearable device. However, existing ECG signal-based methods mainly use features (i.e. frequency domain, time domain, and other nonlinear features) acquired from ECG and its derived signals in order to construct the model. This requires researchers to have rich experience in ECG, which is not common. A convolutional neural network (CNN) is a kind of deep neural network that can automatically learn effective feature representation from training data and has been successfully applied in many fields. Meanwhile, most studies have not considered the impact of adjacent segments on SA detection. Therefore, in this study, we propose a modified LeNet-5 convolutional neural network with adjacent segments for SA detection. Our experimental results show that our proposed method is useful for SA detection, and achieves better or comparable results when compared with traditional machine learning methods.

Detection of Sleep Apnea from Single-Lead ECG Signal Using a Time Window Artificial Neural Network.

Sleep apnea (SA) is a ubiquitous sleep-related respiratory disease. It can occur hundreds of times at night, and its long-term occurrences can lead to some serious cardiovascular and neurological diseases. Polysomnography (PSG) is a commonly used diagnostic device for SA. But it requires suspected patients to sleep in the lab for one to two nights and records about 16 signals through expert monitoring. The complex processes hinder the widespread implementation of PSG in public health applications. Recently, some researchers have proposed using a single-lead ECG signal for SA detection. These methods are based on the hypothesis that the SA relies only on the current ECG signal segment. However, SA has time dependence; that is, the SA of the ECG segment at the previous moment has an impact on the current SA diagnosis. In this study, we develop a time window artificial neural network that can take advantage of the time dependence between ECG signal segments and does not require any prior assumptions about the distribution of training data. By verifying on a real ECG signal dataset, the performance of our method has been significantly improved compared to traditional non-time window machine learning methods as well as previous works.

Simultaneous Ship Detection and Orientation Estimation in SAR Images Based on Attention Module and Angle Regression.

Ship detection and angle estimation in SAR images play an important role in marine surveillance. Previous works have detected ships first and estimated their orientations second. This is time-consuming and tedious. In order to solve the problems above, we attempt to combine these two tasks using a convolutional neural network so that ships may be detected and their orientations estimated simultaneously. The proposed method is based on the original SSD (Single Shot Detector), but using a rotatable bounding box. This method can learn and predict the class, location, and angle information of ships using only one forward computation. The generated oriented bounding box is much tighter than the traditional bounding box and is robust to background disturbances. We develop a semantic aggregation method which fuses features in a top-down way. This method can provide abundant location and semantic information, which is helpful for classification and location. We adopt the attention module for the six prediction layers. It can adaptively select meaningful features and neglect weak ones. This is helpful for detecting small ships. Multi-orientation anchors are designed with different sizes, aspect ratios, and orientations. These can consider both speed and accuracy. Angular regression is embedded into the existing bounding box regression module, and thus the angle prediction is output with the position and score, without requiring too many extra computations. The loss function with angular regression is used for optimizing the model. AAP (average angle precision) is used for evaluating the performance. The experiments on the dataset demonstrate the effectiveness of our method.

[Research on night visibility inversion algorithm based on the dual light sources image and curve evolution theory].

In the present paper, we propose a novel night visibility inversion algorithm. The key idea is based on the curve evolution theory in the context of dual light sources. First, we describe the features of the dual light sources night visibility inversion method, and explain the relationship between fluctuation of light source intensity and curve evolution with the special characters. Then, level set method is introduced ito define the light source intensity and establish the model with visibility inversion, and the algorithm formula is detailed. Experiments results show that, the algorithm is characterized with a strong robustness and compared with the standard visibility, the correlation of visibility result reaches up to 0.98 in the range of 2,000 meter to 12,000 me- ter, which can be used in the visibility monitoring with a large range, high precision and strong robustness.

[Analysis of three-dimensional fluorescence overlapping spectra using nonnegative matrix factorization].

The present paper primarily tests and verifies the effect of NMF in blind source separation of three-dimensional simulative fluorescence spectra, and then four different computational algorithms (multiplicative iterative; alternating least square; second order method; projected gradient algorithm) were used in three practical phenolic compounds (cresol, phenol, thymol) overlapping fluorescence spectra to find out which nonnegatively constrained algorithms is the most efficient for fluorescence spectra unmixing. The experiments demonstrate that four ways have the normalized residuals below 0.06%, and alternating least square (ALS) is the best at both convergence behavior and robustness.

[Research on night visibility estimation method based on image features of dual light sources].

Haze, rain and snow bring a lot of inconvenience in our daily life, especially produce serious potential safety hazard for night transport. In the present paper the authors propose the vision-based dual light sources visibility method to estimate night visibility. This method is significantly advantaged with wide range, high precision and low cost, and has a good robustness in many kinds of weather conditions. Firstly, the authors give the basic visibility estimation model under the atmosphere multiple scattering theory. Secondly, the authors propose the dual light sources method to remove the luminance fluctuations of light sources and the atmosphere light effect, and formulize the algorithm to accurately gain information of light sources from the dual light sources image. At last, the authors design the dual light sources system and conduct a long time experiments under various atmosphere conditions. The experiments show that, with the baseline of 35 m, the visibility range is up to 15 000 m, and relative error is below 20%. This method and system can satisfy the demand of meteorological department and transport agency.

Discrimination of three dimensional fluorescence spectra based on wavelet analysis and independent component analysis.

Fluorescence spectroscopy is a rapid and non-destructive method for monitoring water quality. In this work, wavelet analysis, together with independent component analysis (ICA), was applied for component recognition of seriously overlapped, multi-component, three dimensional fluorescence spectra. Wavelet analysis extracts the features of the spectra and amplifies differences among phenolic homologs. ICA analysis in blind signal separation was used to separate single component before multiple linear regression (MLR). The proposed method increases the correct classification rate and enriches the spectra library. As such, it is a useful alternative to traditional techniques in component recognition.

Insuficiencia velofaríngea en cirugía adenoamigdalina / Velopharyngeal insufficiency in adenotonsillar surgery

Introducción: Insuficiencia velofaríngea (IVF) es una posible complicación asociada a cirugía adenoamigdalina, cuya incidencia real es difícil de establecer, según la literatura, su frecuencia estimada es de 1 en 1.500-10.000 adenoidectomías. Sin embargo, no hay registro en la literatura de IVF transitoria posoperatoria. Nuestra hipótesis del trabajo: la IVF posoperatoria, tanto transitoria como definitiva, está subdiagnosticada o subregistrada. Objetivo: Evaluar la frecuencia y estudiar los factores predisponentes de IVF transitoria y definitiva posoperatoria en una población del Área Occidente de la Región Metropolitana de Chile, 2004-2007. Material y método: Estudio descriptivo y retrospectivo, Revisión de fichas clínicas y registros de la Unidad de Fonoaudiología de pacientes <15 años, operados de patología adenoamigdalina en el Servicio de Otorrinolaringología del Hospital San Juan de Dios, 2004-2007. Resultados: Se registraron 18 casos de IVF transitoria posoperatoria que corresponde a 1,2% (n =1.458). La frecuencia de IVF definitiva posoperatoria fue cero. Conclusión: No se logró analizar los posibles factores predisponentes de IVF posoperatoria. El porcentaje de IVF transitoria encontrado constituye una estadística de referencia para la elaboración del consentimiento informado. Es posible que la frecuencia encontrada pueda ser menor a la real debido al inadecuado registro y control posoperatorio. Creemos que es necesario un protocolo de estudio y de seguimiento de IVF posoperatoria en todos los pacientes intervenidos de cirugía adenoamigdalina.

Introduction: Velopharyngeal insufficiency (VPI) is a possible complication associated with adenotonsillar surgery, whose real incidence is difficult to establish, according to the literature, the estimated frequency is 1 in 1500-10000 adenoidectomies. However, there is no record in the literature of transient postoperative VPI. Hypothesis: The postoperative VPI, both transient and permanent, is underdiagnosed and underreported. Aim: To evaluate the frequency and predisposing factors study transient and permanent postoperative VPI in a population of the West Area of the Metropolitan Region of Chile, 2004-2007. Material and method: A retrospective study, clinical records and records from the Audiology Unit of patients <15 years, operated as adenotonsillar pathology in the Department of Otorhinolaryngology, Hospital San Juan de Dios, from 2004 to 2007. Results: There were 18 cases of transient postoperative VPI which corresponds to 1.2% (n =1458). The frequency of permanent postoperative VPI was zero. Conclusion: It was not possible to analyze the possible predisposing factors for postoperative VPI. The percentage of transient found VPI is a statistical reference for the development of informed consent. It is possible that the frequency found to be lower than actual due to inadequate recording and postoperative control. We believe that we need a study protocol and postoperative monitoring in all VPI patients undergoing adenotonsillar surgery.

Estenosis laringotraqueal: experiencia clínica / Laryngo-tracheal stenosis: our experience

Introducción: La estenosis laringotraqueal es una patología de difícil manejo y obtener un resultado que permita, al sujeto que la padece, volver a tener una función adecuada fonorrespiratoria y deglutoria, no siempre es factible de obtener. Objetivo: Presentar una serie clínica de pacientes que tienen una estenosis de la vía aérea superior (VAS) y que fueron intervenidos quirúrgicamente. Como objetivo secundario es evaluar si a técnicas similares hay o no diferencias entre los grupos etarios. Material y método: Presentamos un análisis retrospectivo de los pacientes intervenidos quirúrgicamente por los autores. En él se realiza la descripción demográfica de los casos, sitio de la estenosis, tipo de intervención según edad; porcentaje de éxito en la decanulación después de una o varias intervenciones, necesidad de reoperación, tipo de injerto y tutores utilizados. Se dividió el grupo en pediátrico (hasta los 18 años) y adulto (mayores de 19 años). Los resultados fueron analizados con las pruebas no paramétricas de x² y de Fisher. Resultados: La casuística consta de 88 pacientes intervenidos quirúrgicamente para reparar una estenosis laringotraqueal. Los menores de 18 años corresponden a 45 casos (51 por ciento). El porcentaje de éxito alcanzado en la primera cirugía es de 75,6 por ciento (34/45 casos) en los menores de 18 años y de 76,7 por ciento por ciento (33/43 casos) en los mayores de 19 años. Se reoperan 15/21 casos fracasados en primera instancia; logrando decanular a 13 de ellos aumentando el éxito a 90,9° por ciento. Fracasan a las reoperaciones y pendientes de resolver aún, 8 casos. La técnica más utilizada fue la reconstrucción laringotraqueal con injerto de cartílago costal ya sea anterior y/o posterior, por ser la estenosis subglótica el sitio de la lesión. El porcentaje de éxito para esta técnica es de 68,3 por ciento; siendo en los menores de...

Introduction: Laryngo-tracheal stenosis is a condition difficult to manage and obtain results which permit the person who suffers it recover phonorespiratory and deglutory function. This is not always possible to achieve. Aim: Present a clinical series of patients with stenosis of the VAS and underwent surgery. A secondary objective is to assess whether or not there are similar technical differences between age groups. Material and method: We present a retrospective analysis of patients surgically treated by the authors. They present demographic description cases, site of stenosis, type of intervention according to age; percentage of successful decannulation after one or more interventions; reoperation, type of graft and stents used. The group was divided into pediatric and adult. Statistical analysis was performed with x2 and Fisher. Results: The case mix consists of 88 patients who underwent surgery to repair laryngo-tracheal stenosis. Children under 18 years correspond to 45 cases (51 percent percent). The percentage of success in the first surgery is 75.6 percent percent (34/45 cases) in children under 18 and 76.7 percent percent (33/43 cases) over 19years. 15/21 failed cases were reoperated in the first instance; 13 of them were decannulated increasing success to 90.9 percent percent. Reoperation failure and still unresolved, 8 cases. Surgical technique used was laryngotracheal reconstruction with costal cartilage graft either anterior or posterior being subglottic the site of stenosis. Success rate to this technique is 68.3 percent percent. In patients under 18 years old is 71 percent percent success and over 18 years 60 percent percent. For tracheal stenosis, tracheal resection with end to end anastomosis has a success rate of over 90 percent and it is performed mostly in the age group over 18 years...