No abstract

No abstract

The lesion regions of a medical image account for only a small part of the image, and a critical imbalance exists in the distribution of the positive and negative samples, which affects the segmentation performance of the lesion regions. Dice loss is beneficial for the image segmentation involving an extreme imbalance of the positive and negative samples but it ignores the background regions, which also contain a large amount of information. In this work, we propose an improved dice loss that can mine the information in background areas and modify network architecture to improve performance. The improved dice loss called weighted soft dice loss (WSDice loss). Our loss function gives a small weight to the background area of the label, so the background area will be added to the calculation when calculating dice loss. It can also soft the hard label in the lesion area to increase the robustness of the model to noise label. What's more, we propose to cascade Focal loss and WSDice loss. Focal Loss is a Distribution-based loss function, WSDice Loss is a Regionbased loss function, the optimization directions of them are different. The cascaded loss function can make full use of the advantages of both and greatly improve model performance. In addition, we add a simple but effective channel attention module to the decode module of U-net. We experimented on the ChestX-ray8 datasets. Compared with Dice loss, WSDice loss improves the dice coefficient by 1.59%, cascaded loss function can improve dice coefficient by 7.81%. The improved in model architecture can increase the dice coefficient by 1.36%.

No abstract

Electrocardiogram (ECG) has been proved to be the most common and effective approach to investigate the cardiovascular disease because that it is simple, non-invasive and low cost. ECG signal automatic classification is a popular research topic and some efficient research work has been done on it. Most of current research work focuses on single ECG label classification, i.e. one ECG signal record corresponds to one label. In practice, one ECG signal usually embraces several cardiovascular diseases at the same time. It is more important to study multi-label ECG signal classification. To our knowledge, few research works have been done on the research topic. To resolve the multi-label ECG signal classification problems, we propose a novel ensemble multi-label classification model in this paper. The model combines several multi-label classification methods to generate a high performance classifier. Mutual information is used to measure the weight of each classifier. At last the ensemble multi-label classification model is used to analyze a clinic ECG signal dataset. The analysis results show that the overall classification performance is improved. It provides a feasible analysis method for multi-label ECG signal automatic classification.