| Summary: | 博士 === 中原大學 === 電子工程研究所 === 92 === In this dissertation, transmitting real-time ECG (electrocardiogram) signals over the mobile telecardiology testbed based on a 3G cellular phone system is considered as a typical example of the next-generation mobile telemedicine applications and its major problem is the transmission errors from a mobile channel. Thus, we propose a scheme of the integration design and a novel control mechanism called medical quality control (mQC) for this testbed to combat its quality degradation caused by the transmission errors. In the scheme, our 3G-based testbed integrates several advanced functions, including SPIHT (Set Partitioning In Hierarchical Tree) compression, UEP (Unequal Error Protection), smart antenna (SA) beamforming, and our mQC mechanisms. The high coding efficiency of SPIHT enables much more efficient use of the bandwidth resource and even saves the 99% bandwidth for the transmission of other medical data/signals. The UEP for SPIHT-based ECG data compression method can combat the interference on the significant information that makes the reconstructed ECG waveform beyond recognition. Smart antenna beamforming could produce a lower bit error rate at a specific SNR than an omnidirectional antenna. Thus, after smart antenna beamforming, a better quality of ECG in terms of lower PRD (Percent of Root mean square Difference) can be obtained under the integration scheme of the testbed.
In addition, mQC is defined to provide the coordination capability among multiple processing tasks over the platforms and the automatic operation based on an objective quality criterion. Hence, two mQC mechanisms, mQC-I and mQC-II, are proposed to operate over this testbed. The mQC-I coordinates two processing tasks, such as SPIHT and UEP, and mQC-II covers the processing tasks in mQC-I along with the function of smart antenna control. A critical design for both is the pilot signal, which reflects the quality degradation caused by transmission errors. The mQC-I enables the automatic and dynamical switching between SPIHT and SPIHT+UEP for ECG test signals, according to the PRD comparison result from the pilot signal. As a result, the PRD obtained with mQC-I is close to the lower one obtained with SPIHT and SPIHT+UEP, regardless of the different degrees of transmission errors. The mQC-II enables beamforming at the Rx only when the PRD of the reconstructed pilot signal is above a specific threshold. The PRD obtained with mQC-II is very close to the case when SA beamforming is always on under the integration scheme of the testbed. Moreover, compared to SPIHT, the mQC-I reduces the amount of the pilot bit stream to maintain the mQC processing gain at higher BER, and improve the quality degradation caused by the introduction of the pilot redundancy bits at lower BER.
Following these cases, the same ideas of two proposed design methodologies can be extended to other platforms and medical signals/data of new or existing mobile telemedicine applications such that they may become the essential features of the next-generation mobile telemedicine platforms in the future.
|
|---|
