Heart failure classification using deep learning to extract spatiotemporal features from ECG.

BACKGROUND: Heart failure is a syndrome with complex clinical manifestations. Due to increasing population aging, heart failure has become a major medical problem worldwide. In this study, we used the MIMIC-III public database to extract the temporal and spatial characteristics of electrocardiogram (ECG) signals from patients with heart failure. METHODS: We developed a NYHA functional classification model for heart failure based on a deep learning method. We introduced an integrating attention mechanism based on the CNN-LSTM-SE model, segmenting the ECG signal into 2 to 20 s long segments. Ablation experiments showed that the 12 s ECG signal segments could be used with the proposed deep learning model for superior classification of heart failure. RESULTS: The accuracy, positive predictive value, sensitivity, and specificity of the NYHA functional classification method were 99.09, 98.9855, 99.033, and 99.649%, respectively. CONCLUSIONS: The comprehensive performance of this model exceeds similar methods and can be used to assist in clinical medical diagnoses.

Circulating microRNAs as biomarkers for severe coronary artery disease.

Coronary artery disease (CAD) is the second leading cause of death after stroke in China. Percutaneous coronary intervention (PCI) significantly improves the prognosis of CAD patients. This study aimed to evaluate the diagnostic value of circulating microRNAs (miRNAs) in patients with severe CAD requiring PCI. The plasma miRNA profiles were determined using miRNA microarray. The relative expression levels of differentially expressed miRNA were measured by quantitative reverse transcription polymerase chain reaction (qRT-PCR). Nine miRNAs (ebv-miR-BART12, ebv-miR-BART16, let-7i-5p, miR-130a-3p, miR-26a-5p, miR-3149, miR-3152-3p, miR-32-3p, and miR-149-3p) were differentially expressed between severe CAD and control groups. Four miRNAs (let-7i-5p, miR-32-3p, miR-3149, and miR-26a-5p) validated by qRT-PCR showed good diagnostic accuracy, with the area under the receiver operating characteristic curves (AUCs) of 0.634 (95% confidence interval [CI] 0.528-0.739), 0.745 (95%CI 0.649-0.84), 0.795 (95%CI 0.709-0.88), and 0.818 (95%CI 0.739-0.897), respectively. Furthermore, the combination of these 4 miRNAs exhibited better diagnostic performance compared with any individual miRNA, with an AUC of 0.837 (95%CI 0.763-0.911). These data indicate that plasma let-7i-5p, miR-32-3p, miR-3149, and miR-26a-5p have promising diagnostic value for severe CAD.

Combination of serum lipids and cancer antigens as a novel marker for colon cancer diagnosis.

BACKGROUND: Colon cancer is a malignancy of the large intestine with high mortality and economic burden. Recent studies reveal a new relationship between blood lipids and the risk of cancer. The presents study aims to investigate the combination of serum lipids with cancer antigens as a novel diagnostic marker for colon cancer. METHODS: Two hundred of colon cancer patients or healthy subjects were recruited. Serum lipids and cancer antigens such as total cholesterol (TC), high-density lipoprotein (HDL), carcinoembryonic antigen (CEA) and carbohydrate antigen 19-9 (CA19-9) were measured. RESULTS: There were significantly lower level of serum TC or HDL, and significantly higher level of serum CEA or CA19-9 in patients than in healthy subjects. Serum TC or HDL in patients with advanced colon cancer was significantly lower than the ones with early stage disease. The level of serum TC or HDL in patients after surgical removal of colon cancer was significantly higher compared to the ones before surgery, but serum CEA or CA19-9 after surgery was significantly reduced in comparison with the ones before surgery. Combined TC, HDL, CEA and CA19-9 as a diagnostic marker for colon cancer had the highest positive predictive rate in comparison with individual, two or three of the parameters. CONCLUSIONS: The combination of serum TC, HDL, CEA and CA19-9 can be used as an effective marker for colon cancer, and offers a novel strategy for clinical diagnosis and monitoring the disease.