Automatic recognition of murmurs of ventricular septal defect using convolutional recurrent neural networks with temporal attentive pooling.

Recognizing specific heart sound patterns is important for the diagnosis of structural heart diseases. However, the correct recognition of heart murmur depends largely on clinical experience. Accurately identifying abnormal heart sound patterns is challenging for young and inexperienced clinicians. This study is aimed at the development of a novel algorithm that can automatically recognize systolic murmurs in patients with ventricular septal defects (VSDs). Heart sounds from 51 subjects with VSDs and 25 subjects without a significant heart malformation were obtained in this study. Subsequently, the soundtracks were divided into different training and testing sets to establish the recognition system and evaluate the performance. The automatic murmur recognition system was based on a novel temporal attentive pooling-convolutional recurrent neural network (TAP-CRNN) model. On analyzing the performance using the test data that comprised 178 VSD heart sounds and 60 normal heart sounds, a sensitivity rate of 96.0% was obtained along with a specificity of 96.7%. When analyzing the heart sounds recorded in the second aortic and tricuspid areas, both the sensitivity and specificity were 100%. We demonstrated that the proposed TAP-CRNN system can accurately recognize the systolic murmurs of VSD patients, showing promising potential for the development of software for classifying the heart murmurs of several other structural heart diseases.

Blind Monaural Source Separation on Heart and Lung Sounds Based on Periodic-Coded Deep Autoencoder.

Auscultation is the most efficient way to diagnose cardiovascular and respiratory diseases. To reach accurate diagnoses, a device must be able to recognize heart and lung sounds from various clinical situations. However, the recorded chest sounds are mixed by heart and lung sounds. Thus, effectively separating these two sounds is critical in the pre-processing stage. Recent advances in machine learning have progressed on monaural source separations, but most of the well-known techniques require paired mixed sounds and individual pure sounds for model training. As the preparation of pure heart and lung sounds is difficult, special designs must be considered to derive effective heart and lung sound separation techniques. In this study, we proposed a novel periodicity-coded deep auto-encoder (PC-DAE) approach to separate mixed heart-lung sounds in an unsupervised manner via the assumption of different periodicities between heart rate and respiration rate. The PC-DAE benefits from deep-learning-based models by extracting representative features and considers the periodicity of heart and lung sounds to carry out the separation. We evaluated PC-DAE on two datasets. The first one includes sounds from the Student Auscultation Manikin (SAM), and the second is prepared by recording chest sounds in real-world conditions. Experimental results indicate that PC-DAE outperforms several well-known separation works in terms of standardized evaluation metrics. Moreover, waveforms and spectrograms demonstrate the effectiveness of PC-DAE compared to existing approaches. It is also confirmed that by using the proposed PC-DAE as a pre-processing stage, the heart sound recognition accuracies can be notably boosted. The experimental results confirmed the effectiveness of PC-DAE and its potential to be used in clinical applications.

S1 and S2 Heart Sound Recognition Using Deep Neural Networks.

OBJECTIVE: This study focuses on the first (S1) and second (S2) heart sound recognition based only on acoustic characteristics; the assumptions of the individual durations of S1 and S2 and time intervals of S1-S2 and S2-S1 are not involved in the recognition process. The main objective is to investigate whether reliable S1 and S2 recognition performance can still be attained under situations where the duration and interval information might not be accessible. METHODS: A deep neural network (DNN) method is proposed for recognizing S1 and S2 heart sounds. In the proposed method, heart sound signals are first converted into a sequence of Mel-frequency cepstral coefficients (MFCCs). The K-means algorithm is applied to cluster MFCC features into two groups to refine their representation and discriminative capability. The refined features are then fed to a DNN classifier to perform S1 and S2 recognition. We conducted experiments using actual heart sound signals recorded using an electronic stethoscope. Precision, recall, F-measure, and accuracy are used as the evaluation metrics. RESULTS: The proposed DNN-based method can achieve high precision, recall, and F-measure scores with more than 91% accuracy rate. CONCLUSION: The DNN classifier provides higher evaluation scores compared with other well-known pattern classification methods. SIGNIFICANCE: The proposed DNN-based method can achieve reliable S1 and S2 recognition performance based on acoustic characteristics without using an ECG reference or incorporating the assumptions of the individual durations of S1 and S2 and time intervals of S1-S2 and S2-S1.

Association of interleukin-16 polymorphisms with graves' disease in a Taiwanese population.

Graves' disease (GD) is a complex, organ-specific autoimmune disease wherein the thyroid gland becomes enlarged and overactive. During GD progression, T cells secrete interleukin-16 (IL-16) to promote inflammation, act as chemoattractants that recruit more inflammatory cells, and activate target cells to enhance the development of GD. To investigate the role of IL-16 in GD, we genotyped 474 patients with GD at 8 single-nucleotide polymorphisms (SNPs) in the IL-16 gene. The IL-16 SNP rs8028364 was found to be associated with GD when compared with the control subjects (P = 2.93 × 10⁻¹7; CG genotype: odds ratio [OR] = 0.2 [0.07, 0.59]; CC genotype: OR = 0.03 [0.01, 0.09]). The rs1131445 polymorphism was found to be associated with GD under the allelic model (P = 0.01; G allele: OR = 1.97 [1.17, 3.32]). Sliding-window haplotype analysis by the PLINK program showed that the most significant haplotype was provided by the 6-SNP haplotype window, consisting of rs7182786, rs8028364, rs12907134, rs4128767, rs4072111 and rs8031107 (P = 2.31 × 10⁻5¹). We found 2 protective haplotypes: GCAAGG (P = 8.69 × 10⁻7; OR = 0.22 [0.12, 0.41]) and AGAAGG (P = 0.0012; OR = 0.26 [0.12, 0.6]). In addition, GGGGAA (P = 0.39; OR = 2.32 [1.08, 4.99]) and GGGAGA (P = 1.18 × 10⁻5; OR = 5.54 [2.50, 12.31]) were found to be the two high-risk haplotypes. These results suggest that polymorphisms in IL-16 may be used as genetic markers for the diagnosis and prognosis of GD.

Apicidin-resistant HA22T hepatocellular carcinoma cells massively promote pro-survival capability via IGF-IR/PI3K/Akt signaling pathway activation.

Despite rapid advances in the diagnostic and surgical procedures, hepatocellular carcinoma (HCC) remains one of the most difficult human malignancies to treat. This may be due to the chemoresistant behaviors of HCC. It is believed that acquired resistance could be overcome and improve the overall survival of HCC patients by understanding the mechanisms of chemoresistance in HCC. A stable HA22T cancer line, which is chronically resistant to a histone deacetylase inhibitor, was established. After comparing the molecular mechanism of apicidin-R HA22T cells to parental ones by Western blotting, cell cycle-regulated proteins did not change in apicidin-R cells, but apicidin-R cells were more proliferative and had higher tumor growth (wound-healing assay and nude mice xenograft model). Moreover, apicidin-R cells displayed increased levels of p-IGF-IR, p-PI3K, p-Akt, Bcl-xL, and Bcl-2 but also significantly inhibited the tumor suppressor PTEN protein and apoptotic pathways when compared to the parental strain. Therefore, the highly proliferative effect of apicidin-R HA22T cells was blocked by Akt knockdown. For all these findings, we believe that novel strategies to attenuate IGF-IR/PI3K/Akt signaling could overcome chemoresistance toward the improvement of overall survival of HCC patients.

Dung-Shen Downregulates the Synergistic Apoptotic Effects of Angiotensin II Plus Leu 27-IGF II on Cardiomyoblasts.

BACKGROUND: Insulin growth factor II (IGFII) is expressed after ischemic stress in pig hearts and after myocardial infarction in humans. However, its receptor (IGFIIR) cannot be found in normal adult hearts. Moreover, a mouse IGFII overexpression model showed a heart and kidney hypertrophy phenomenon similar to Beckwith-Wiedemann syndrome in humans. The previous studies from our lab showed that an increase in AngII in H9c2 cells causes an elevation in IGFII and IGFIIR through MEK and JNK activation, leading to a rise in intracellular Ca(2+) ions, activation of calcineurin by PLC-ß3 via Gαq, insertion into mitochondrial membranes of BAD, and apoptosis via activation of caspases 9 and 3. Codonopsis pilosula (Dung-shen) has various uses in traditional Chinese medicine, including lowering blood pressure, and increasing red and white blood cell counts. METHODS: The purpose of our study is to investigate whether the addition of C. pilosula will attenuate the AngII plus Leu27-IGFII-induced apoptosis in H9c2 cardiomyoblast cells. RESULTS: From MTT [3-(4, 5-dimethyl-2-thiazolyl)-2, 5-diphenyl-2-tetrazolium bromide] results, it was revealed that AngII plus Leu(27)-IGFII significantly reduced cell viability, which was reversed by C. pilosula. Additionally, C. pilosula also reversed apoptosis (TUNEL staining) increased by AngII plus Leu27-IGFII. Up-regulation of caspase 3 by AngII plus Leu27-IGFII was attenuated by C. pilosula treatment, as shown in western blotting assay and immunofluorescence microscopy results. CONCLUSIONS: C. pilosula is able to suppress the apoptotic pathway enhanced by AngII plus Leu27-IGFII in myocardial cells. KEY WORDS: Angiotensin II; Apoptosis; Codonopsis pilosula; Leucine27-insulin like growth factor II; Mitochondrial outer membrane permeability.

The impact of polymorphisms in STAT6 on treatment outcome in HCV infected Taiwanese Chinese.

Genetic polymorphisms observed in various disease states associated with sensitivity or resistance to specific treatments have been a robust area of investigation for decades, with the potential to allow clinicians to make evidence-based decisions on the appropriate course of treatment. This study aimed to evaluate whether genetic polymorphisms of the signal transducer and activator of transcription 6 gene (STAT6) could be associated with a sustained virological response (SVR) among patients infected with hepatitis C virus genotypes 1 and 2 (HCV-1 and HCV-2) who were treated with peginterferon plus ribavirin (PEG-IFNα-RBV). We analyzed the associations between SVR to PEG-IFNα-RBV therapy and 4 single nucleotide polymorphisms (SNPs) in STAT6. This study included Taiwanese Chinese patients infected with either HCV-1 (n = 265) or HCV-2 (n = 195) in the presence or absence of an SVR. Among the STAT6 SNPs examined, the dosage effect of the A allele and allele frequency in rs1059513 were inversely correlated with SVR in patients infected with HCV-1 (P = 0.0179 and P = 0.0235, respectively). This effect was not observed in patients infected with HCV-2. The GG, GGG, and GGGC STAT6 haplotypes comprising 2, 3, and 4 SNPs (rs1059513, rs703817, rs324015, and rs3024974) were found to be associated with SVR, and their presence may increase the probability of a successful treatment outcome in patients infected with HCV-1 (P = 0.0273, 0.0352, and 0.0368, respectively). Moreover, a multivariate logistic regression model for predicting an SVR revealed that the presence of the GGGC haplotype carriers mutually affected the outcome of PEG-IFNα-RBV treatment. The presence of STAT6 SNPs and the association with SVR demonstrated that STAT6 polymorphisms might influence the therapeutic outcomes of patients infected with HCV-1 under standard-of-care (SOC) treatment.

Dung-shen (Codonopsis pilosula) attenuated the cardiac-impaired insulin-like growth factor II receptor pathway on myocardial cells.

Previous studies from our lab showed that increase in AngII in H9c2 cells causes elevated IGFII and IGFIIR through MEK and JNK, leading to rise in intracellular calcium, calcineurin activation by PLC-ß3 via Gαq, insertion into mitochondrial membranes of Bad, and apoptosis via caspases 9 and 3. Codonopsis pilosula is traditionally used to lower blood pressure. The purpose of our study is to investigate if C. pilosula attenuates AngII plus Leu(27)-IGFII-induced calcium influx and apoptosis in H9c2 cardiomyoblasts. C. pilosula significantly attenuated AngII induced IGFIIR promoter activity. Leu(27)-IGFII was applied to enhance the AngII effect. C. pilosula also reversed Ca(2+) influx, MOMP and apoptosis increased by AngII plus Leu(27)-IGFII. Molecular markers in IGFIIR apoptotic pathway (IGFIIR, calcineurin, etc.) and IGFIIR-Gαq association were downregulated by C. pilosula. However, p-Bad(Ser136) and Bcl-2 were increased. Therefore, C. pilosula suppresses AngII plus Leu(27)-IGFII-induced IGFII/IGFIIR pathway in myocardial cells.

Rhubarb inhibits hepatocellular carcinoma cell metastasis via GSK-3-ß activation to enhance protein degradation and attenuate nuclear translocation of ß-catenin.

The aim of our study was to investigate the mechanisms by which rhubarb regulates ß-catenin as well as metastasis of hepatocellular carcinomas. Our results revealed that rhubarb extract inhibited HA22T cell migration ability in wound healing, migration and invasion assays in a dose-dependent manner. Rhubarb also reduced ß-catenin protein level, downregulated its downstream proteins, cyclin D, Tbx3 and c-Myc, and attenuated the expression of MMP9 and contactin-1 metastatic factors. Additionally, rhubarb inhibited ß-catenin nuclear accumulation and induced its degradation via proteasome-mediated pathway. Furthermore, we found that rhubarb suppressed the p-ser(9) GSK-3-ß protein level to inactivate Wnt signalling and reduce ß-catenin protein level. Taken together; we found that rhubarb blocked the metastatic process of HA22T hepatocellular carcinoma cells mediated through GSK-3-ß activation, and enhancement of protein degradation as well as reduction of the nuclear accumulation of ß-catenin.

Thymic stromal lymphopoietin gene promoter polymorphisms and expression levels in Graves' disease and Graves' ophthalmopathy.

BACKGROUND: Graves disease (GD) is an organ-specific autoimmune disease characterized by hyperthyroidism, diffuse goiter, autoantibodies against thyroid-specific antigens, and dermopathy. Studies of GD have demonstrated the importance of the Th2 and Th17 immune responses in mediating disease progression. In the present study, we investigated the role of a Th2 cytokine, thymic stromal lymphopoietin (TSLP), in GD and Th17 differentiation. METHODS: In this study, we genotyped 470 patients with GD at 3 single nucleotide polymorphisms (SNPs) in TSLP. In addition, the serum concentrations of TSLP were determined in 432 patients and 272 controls. Ten patients and controls each were further screened using in vitro Th17 differentiation assays. The SNPs were genotyped using ABI TaqMan® SNP genotyping assays. For the Th17 differentiation assays, peripheral blood mononuclear cells (PBMCs) isolated from the patients and controls were placed into Th17 differentiation media, and interleukin 17 expression levels were determined. RESULTS: Haplotype analysis indicated that patients with the Ht3 (TCC) haplotype have a 3.28-fold higher risk of developing GD (p = 0.007), whereas those with the Ht5 (TCG) haplotype had a 0.03-fold, reduced risk of developing GD (p = 1 × 10-14). SNP rs3806933 (p = 0.007) was associated with female Graves ophthalmopathy (GO). TSLP expression levels were higher in GD patients than in control subjects, and TLSP was also shown to promote the differentiation of Th17 cells in GD patients. CONCLUSIONS: These results suggest that polymorphisms in TSLP may be used as genetic markers for the diagnosis and prognosis of GD. Furthermore, TLSP may be a target for treating GD.

NADPH oxidase-derived superoxide anion-induced apoptosis is mediated via the JNK-dependent activation of NF-κB in cardiomyocytes exposed to high glucose.

Hyperglycemia-induced generation of reactive oxygen species (ROS) can lead to cardiomyocyte apoptosis and cardiac dysfunction. However, the mechanism by which high glucose causes cardiomyocyte apoptosis is not clear. In this study, we investigated the signaling pathways involved in NADPH oxidase-derived ROS-induced apoptosis in cardiomyocytes under hyperglycemic conditions. H9c2 cells were treated with 5.5 or 33 mM glucose for 36 h. We found that 33 mM glucose resulted in a time-dependent increase in ROS generation as well as a time-dependent increase in protein expression of p22(phox), p47(phox), gp91(phox), phosphorylated IκB, c-Jun N-terminal kinase (JNK) and p38, as well as the nuclear translocation of NF-kB. Treatment with apocynin or diphenylene iodonium (DPI), NADPH oxidase inhibitors, resulted in reduced expression of p22(phox), p47(phox), gp91(phox), phosphorylated IκB, c-Jun N-terminal kinase (JNK) and p38. In addition, treatment with JNK and NF-kB siRNAs blocked the activity of caspase-3. Furthermore, treatment with JNK, but not p38, siRNA inhibited the glucose-induced activation of NF-κB. Similar results were obtained in neonatal cardiomyocytes exposed to high glucose concentrations. Therefore, we propose that NADPH oxidase-derived ROS-induced apoptosis is mediated via the JNK-dependent activation of NF-κB in cardiomyocytes exposed to high glucose.