Article Preview
Top1. Introduction
Electronic ubiquitous healthcare (Varshney, 2003; Hu, Stoelting, Wang, Zou, & Sarrafzadeh, 2010) has undergone enormous advancement thanks to recent Radio Frequency (RF) microelectronics technologies and low power computing devices, such as body area sensor networks, Electrocardiagram (ECG) monitoring on mobile platforms, etc. In addition, due to the upcoming retirement of the baby boomer generation, wireless ubiquitous elder healthcare based on advanced information technologies shows considerable markets. In general, although doctors and nurses may be hundreds miles away from the patients, wireless ubiquitous tele-medicine should provide high quality biomedical signals equivalent to on-site measurement in hospitals.
However, due to error-prone nature, wireless channel can only provide limited Quality of Service (QoS) for healthcare data flows which poses significant challenge of high precision healthcare guarantee. Furthermore, healthcare devices such as body sensors and cardio data transceivers have limited circuit board sizes, constrained microprocessor/microcontroller computing capability, and low power low cost cheap radio chips providing limited channel bandwidth.
Figure 1 illustrates such an example of electronic healthcare using wireless mobile networks and internets. The patients (for example, cardio patients) are liberated from wires and hospitals by carrying an ECG wireless sensor PDA based health monitoring device. We also illustrate a real-world ECG sensing device developed by our research teams. Such devices acquire cardio information in real-time and transmits the physiological information via wireless networks to home computers, and the home computers relay the ECG information through the internet backbones to hospital servers. The patients’ cardio information will be presented on the screen in front of the physicians, doctors or nurses hundreds or thousands of miles away from the patients. In such systems, the bandwidth constraints of the wireless networks and the ECG signal quality are both critical challenges impacting the system performance.
Figure 1. Illustration of an electronic healthcare system with wireless ECG monitoring
To deal with such challenge, we propose a new Region of Interest (ROI) based approach to improve healthcare data service quality with limited bandwidth resource. Specifically, important bit sequence in healthcare data is extracted once it falls in the ROI. Then variable bit-plane coding is applied to bit sequence, treating ROI and non-ROI unequally. More bit-planes are used to code the fine-detailed ROI information to improve precision, and less bit-planes are applied to non-ROI information to reduce total data rate consumption. For convenience, we utilize ECG as an example of biometric data in the wireless cardio health monitoring scenarios.
This paper is organized as follows. In Section 1 we have briefly introduced the background information. In Section 2 we review related works and pointed out the uniqueness of this research. In Section 3 we formulate the mathematical problem as a quality optimization problem with rate constraint. In Section 4 we analyze the problem and provide solutions. In Section 5 we perform numerical simulation study and in Section 6 we draw conclusion.