1 September 2020 in Journal of Physics: Conference Series
Congestive heart failure (CHF) is a common serious heart disease that requires a number of clinical examinations to diagnose, which are costly and time-consuming. Electrocardiogram (ECG) is widely used in the diagnosis of various cardiovascular diseases due to its advantages of non-invasive, convenient and cheap, so the automatic CHF detection algorithm based on ECG signals can overcome the above shortcomings and has great application prospects. In this paper, inspired by the idea of DenseNet in computer version, we refined it to be applicable to CHF detection task, thus improving the diagnosis accuracy of the model. Secondly, to improve the robustness of the algorithm, we built a CHF database on PhysioBank, which contained more diverse data compared with similar studies, and conducted experiments on the built database. Finally, we presented an evaluation method based on the “inter-patient” pattern to evaluate the performance of the method more objectively. The results show that our algorithm can efficiently detect CHF with accuracy, sensitivity and specificity up to 94.97%, 89.38% and 99.50%, respectively. The algorithm proposed in this study can provide reliable references for doctors, and can be used in portable devices to realize real-time monitoring for patients.
22 November 2018 in Mobile Information Systems
Correctly identifying human activities is very significant in modern life. Almost all feature extraction methods are based directly on acceleration and angular velocity. However, we found that some activities have no difference in acceleration and angular velocity. Therefore, we believe that for these activities, any feature extraction method based on acceleration and angular velocity is difficult to achieve good results. After analyzing the difference of these indistinguishable movements, we propose several new features to improve accuracy of recognition. We compare the traditional features and our custom features. In addition, we examined whether the time-domain features and frequency-domain features based on acceleration and angular velocity are different. The results show that (1) our custom features significantly improve the precision of the activities that have no difference in acceleration and angular velocity; and (2) the combination of time-domain features and frequency-domain features does not significantly improve the recognition of different activities.