7 February 2022 in IEEE Transactions on Neural Networks and Learning Systems
Artificial intelligence and machine learning techniques have progressed dramatically and become powerful tools required to solve complicated tasks, such as computer vision, speech recognition, and natural language processing. Since these techniques have provided promising and evident results in these fields, they emerged as valuable methods for applications in human physiology and healthcare. General physiological recordings are time-related expressions of bodily processes associated with health or morbidity. Sequence classification, anomaly detection, decision making, and future status prediction drive the learning algorithms to focus on the temporal pattern and model the nonstationary dynamics of the human body. These practical requirements give birth to the use of recurrent neural networks (RNNs), which offer a tractable solution in dealing with physiological time series and provide a way to understand complex time variations and dependencies. The primary objective of this article is to provide an overview of current applications of RNNs in the area of human physiology for automated prediction and diagnosis within different fields. Finally, we highlight some pathways of future RNN developments for human physiology.
10 July 2019 in Royal Society Open Science
Hyoid bone movement is an important physiological event during swallowing that contributes to normal swallowing function. In order to determine the adequate hyoid bone movement, clinicians conduct an X-ray videofluoroscopic swallowing study, which even though it is the gold-standard technique, has limitations such as radiation exposure and cost. Here, we demonstrated the ability to track the hyoid bone movement using a non-invasive accelerometry sensor attached to the surface of the human neck. Specifically, deep neural networks were used to mathematically describe the relationship between hyoid bone movement and sensor signals. Training and validation of the system were conducted on a dataset of 400 swallows from 114 patients. Our experiments indicated the computer-aided hyoid bone movement prediction has a promising performance when compared with human experts’ judgements, revealing that the universal pattern of the hyoid bone movement is acquirable by the highly nonlinear algorithm. Such a sensor-supported strategy offers an alternative and widely available method for online hyoid bone movement tracking without any radiation side-effects and provides a pronounced and flexible approach for identifying dysphagia and other swallowing disorders.
1 June 2020 in Journal of Power Sources
1 October 2017 in Measurement
20 August 2021 in IEEE Journal of Biomedical and Health Informatics
Aspiration is a serious complication of swallowing disorders. Adequate detection of aspiration is essential in dysphagia management and treatment. High-resolution cervical auscultation has been increasingly considered as a promising noninvasive swallowing screening tool and has inspired automatic diagnosis with advanced algorithms. The performance of such algorithms relies heavily on the amount of training data. However, the practical collection of cervical auscultation signal is an expensive and time-consuming process because of the clinical settings and trained experts needed for acquisition and interpretations. Furthermore, the relatively infrequent incidence of severe airway invasion during swallowing studies constrains the performance of machine learning models. Here, we produced supplementary training exemplars for desired class by capturing the underlying distribution of original cervical auscultation signal features using auxiliary classifier Wasserstein generative adversarial networks. A 10-fold subject cross-validation was conducted on 2079 sets of 36-dimensional signal features collected from 189 patients undergoing swallowing examinations. The proposed data augmentation outperforms basic data sampling, cost-sensitive learning and other generative models with significant enhancement. This demonstrates the remarkable potential of proposed network in improving classification performance using cervical auscultation signals and paves the way of developing accurate noninvasive swallowing evaluation in dysphagia care.
1 June 2016 in IET Power Electronics
This study investigates the near‐field wireless power transfer system with small‐sized receiver. The receiver is placed at a relatively long distance from the power emitter. The coupling between the receiver and the emitter is very weak, which results in low‐power delivery and efficiency. A passive resonance coil, which enhances the field, is added to the emitter to address these problems. The necessary coupled‐mode theory is developed and combined with optimal operation of the class‐E amplifier. The impact of the enhancing coil (EX) on the energy distribution is discussed from the perspective of energy amplification between the source coil and the EX. The point of the maximum energy of the EX is found and necessary conditions for improving transfer efficiency of such structure are proposed. An experimental setup operating at a frequency of 300 kHz was built using the EX concept. The maximum secondary magnetic field generated by the EX was established and used to wirelessly power a 4 cm receiver with more than 40 mW at 0.5 m. The structure is also compared with a similar system without an EX. Test results show that the inclusion of the EX in the emitter strengthens the spatial magnetic field and improves the transmission efficiency.