English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Plus annual

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Plus monthly

Global: Zendy Tools annual

Global: Zendy Tools monthly

Global: Zendy Free Plan

Abnormal heart sounds detection is of great value for the pre-diagnosis of heart disease. Since the pathological information is usually contained in each heart cycle, it is essential to segment the heart sounds before further analysis. In order to extract more detailed information, the time and frequency domain analysis of the heart cycle is necessary. So the time-frequency representation (TFR) of each heart cycle is extracted in this study. However, the durations of the heart cycles between different samples are usually not the same. As a result, the sizes of the corresponding TFR are different which prohibits the direct comparison between them. To solve this problem, the TFRs are scaled to a fixed size through the bilinear interpolation method. Nevertheless, the scaled TFR contains some noises which are useless for classification. For the purpose of removing noises, a feature selection method based on similarity is applied. Then, the selected features are input to a support vector machine (SVM) for classification. At last, the proposed method is evaluated on the dataset offered by the PhysioNet/Computing in Cardiology Challenge 2016. The overall score 85.40% is achieved by a 5-fold cross-validation. It is the average of the sensitivity(75.03%) and specificity(95.76%). The best overall performance of our method on the challenge is 84%.

Abnormal Heart Sounds detection based on the Scaled Time:Frequency Representation and Feature Selection