Detection of Coronary Artery Disease Based on Clinical Phonocardiogram and Multiscale Attention Convolutional Compression Network.

Heart sound is an important physiological signal that contains rich pathological information related with coronary stenosis. Thus, some machine learning methods are developed to detect coronary artery disease (CAD) based on phonocardiogram (PCG). However, current methods lack sufficient clinical dataset and fail to achieve efficient feature utilization. Besides, the methods require complex processing steps including empirical feature extraction and classifier design. To achieve efficient CAD detection, we propose the multiscale attention convolutional compression network (MACCN) based on clinical PCG dataset. Firstly, PCG dataset including 102 CAD subjects and 82 non-CAD subjects was established. Then, a multiscale convolution structure was developed to catch comprehensive heart sound features and a channel attention module was developed to enhance key features in multiscale attention convolutional block (MACB). Finally, a separate downsampling block was proposed to reduce feature losses. MACCN combining the blocks can automatically extract features without empirical and manual feature selection. It obtains good classification results with accuracy 93.43%, sensitivity 93.44%, precision 93.48%, and F1 score 93.42%. The study implies that MACCN performs effective PCG feature mining aiming for CAD detection. Further, it integrates feature extraction and classification and provides a simplified PCG processing case.

Geographical traceability of soybean: An electronic nose coupled with an effective deep learning method.

The quality of soybeans is correlated with their geographical origin. It is a common phenomenon to replace low-quality soybeans from substandard origins with superior ones. This paper proposes the adaptive convolutional kernel channel attention network (AKCA-Net) combined with an electronic nose (e-nose) to achieve soybean quality traceability. First, the e-nose system is used to collect soybean gas information from different origins. Second, depending on the characteristics of the gas information, we propose the adaptive convolutional kernel channel attention (AKCA) module, which focuses on key gas channel features adaptively. Finally, the AKCA-Net is proposed, which is capable of modeling deep gas channel interdependency efficiently, realizing high-precision recognition of soybean quality. In comparative experiments with other attention mechanisms, AKCA-Net demonstrated superior performance, achieving an accuracy of 98.21%, precision of 98.57%, and recall of 98.60%. In conclusion, the combination of the AKCA-Net and e-nose provides an effective strategy for soybean quality traceability.

AUNet: a deep learning method for spectral information classification to identify inks.

It is common to tamper with the contents of documents and forge contracts illegally. In this work, we propose a U-shaped network with attention modules (AUNet) and combine it with a hyperspectral system to effectively identify different inks. It provides an effective detection method for illegal tampering with documents and forging contract contents. First, the hyperspectral system obtains the spectral information of different pen inks without destroying the sample. Second, because the hyperspectral system's detection data have the characteristics of small samples, we introduce U-Net to conduct the deep fusion of multi-level spectral information to avoid feature degradation and fully mine the deep features hidden in the spectral information. Finally, spatial and channel attention modules are introduced to focus on the features affecting classification performance. The results show that AUNet effectively realizes the effective classification of ink spectral information and achieves 97.81% accuracy, 98.71% recall, 98.80% precision, and 98.71% F1-score.

A residual dense comprehensively regulated convolutional neural network to identify spectral information for egg quality traceability.

In the egg market, due to the different nutritional values of eggs laid by hens under different feeding conditions, it is common for low-quality eggs to be counterfeited as high-quality eggs. This paper proposes a residual dense comprehensively regulated convolutional neural network (RDCR-Net) to identify the quality of eggs laid by hens under different feeding conditions. Firstly, a hyperspectral system is used to obtain the spectral information of eggs. Secondly, due to the complex structure of the spectral information, a comprehensively regulated convolution (CRConv) is proposed to extract features hidden in the spectral information through feature transformation in multiple spaces. Thirdly, due to the limited availability of spectral information training samples, deep networks may suffer from feature degradation. The residual dense comprehensively regulated block (RDCR-Block) is proposed to tightly connect multiple CRConv layers with residual dense connections. Finally, the RDCR-Block is taken as the central unit, and the RDCR-Net is designed to identify egg spectral information. In the comparison of multi-model results, the RDCR-Net obtains the best classification performance with 96.29% accuracy, 97.53% precision, 97.14% recall, and 96.19% kappa coefficient. In summary, the RDCR-Net effectively extracts the deep features of spectral information, achieves high accuracy in identifying eggs laid by hens under different feeding conditions, and provides a new method for egg quality traceability.