A High Accuracy & Ultra-Low Power ECG-Derived Respiration Estimation Processor for Wearable Respiration Monitoring Sensor.

The respiratory rate is widely used for evaluating a person's health condition. Compared to other invasive and expensive methods, the ECG-derived respiration estimation is a more comfortable and affordable method to obtain the respiration rate. However, the existing ECG-derived respiration estimation methods suffer from low accuracy or high computational complexity. In this work, a high accuracy and ultra-low power ECG-derived respiration estimation processor has been proposed. Several techniques have been proposed to improve the accuracy and reduce the computational complexity (and thus power consumption), including QRS detection using refractory period refreshing and adaptive threshold EDR estimation. Implemented and fabricated using a 55 nm processing technology, the proposed processor achieves a low EDR estimation error of 0.73 on CEBS database and 1.2 on MIT-BIH Polysomnographic Database while demonstrating a record-low power consumption (354 nW) for the respiration monitoring, outperforming the existing designs. The proposed processor can be integrated in a wearable sensor for ultra-low power and high accuracy respiration monitoring.

TSE-CNN: A Two-Stage End-to-End CNN for Human Activity Recognition.

Human activity recognition has been widely used in healthcare applications such as elderly monitoring, exercise supervision, and rehabilitation monitoring. Compared with other approaches, sensor-based wearable human activity recognition is less affected by environmental noise and therefore is promising in providing higher recognition accuracy. However, one of the major issues of existing wearable human activity recognition methods is that although the average recognition accuracy is acceptable, the recognition accuracy for some activities (e.g., ascending stairs and descending stairs) is low, mainly due to relatively less training data and complex behavior pattern for these activities. Another issue is that the recognition accuracy is low when the training data from the test subject are limited, which is a common case in real practice. In addition, the use of neural network leads to large computational complexity and thus high power consumption. To address these issues, we proposed a new human activity recognition method with two-stage end-to-end convolutional neural network and a data augmentation method. Compared with the state-of-the-art methods (including neural network based methods and other methods), the proposed methods achieve significantly improved recognition accuracy and reduced computational complexity.