Detecting Human Physiological RF Signals for Remote Medical Care With LabVIEW and PXI

While the concept of noncontact detection of vital signs was demonstrated before 2000, research efforts in this century have been moving the technology toward reality with advanced hardware and signal processing algorithms.



The Need for a PXI-Based Noncontact Vital Sign Detector



Prior to our research, the Doppler radar technique was used to sense physiological movements. However, these techniques were mostly instrument based and required several advanced RF/microwave components such as a spectrum analyzer, signal generator, and signal analyzer. Thus, the cost of building a bioradar system was high. In addition, once the measurements were taken, post signal processing had to be performed using different techniques. However, by implementing the system using NI PXI hardware, we did not require separate microwave components and we achieved real-time signal processing simultaneously with data acquisition in LabVIEW.



System Requirements and Setup



The PXI-based radar system requires the following instruments:



-NI PXIe-5663 vector signal analyzer, which consists of an NI PXIe-5652 RF signal generator, an NI PXIe-5601 RF downconverter, and an NI PXIe-5622 16-bit IF digitizer

-NI PXIe-5673 vector signal generator, which consists of an NI PXIe-5450 400 MS/s I/Q signal generator, an NI PXIe-5611 I/Q vector modulator, and an NI PXI-5652

-NI PXI-5691 preamplifier

-Two patch antennas, which can be replaced with other types of antennas



The entire PXI instrument is housed in an NI PXIe-1075 chassis and is controlled by an NI PXIe-8133. Other types of chassis and controllers can be used.



Figure 1 (top right) shows the port connections for the entire system. The NI PXIe-5673 VSG generates a 5.8 GHz single-tone signal, which is transmitted by the transmitting antenna toward the subject under test. When the signal reaches the human body, it is bounced back and modulated by physiological motions including respiration and heartbeat. The reflected signal is captured by the receiving antenna (and first amplified by the PXI-5691 preamplifier. After that, the received signal is fed into the NI PXIe-5663 VSA, where it is first downconverted by the NI PXIe-5601, with the NI PXIe-5652 providing a 5.8 GHz local oscillator signal. The IF signal is then digitized by the NI PXIe-5622. Then the signal is split into I and Q channel outputs.



The patch antennas were chosen to transmit and receive signals at the same frequency as the VSG-generated frequency. The output power level was set to 5 dBm in our demonstration, which is the maximum output power level for the VSG.



Note that ideally a single LO should be used for both the transmitter and the receiver to take advantage of the range correlation effect to minimize the oscillator phase noise. However, it is much easier to control the hardware if the VSA and the VSG use two different LOs. To synchronize the VSG with the VSA, the PXI 10 MHz backplane clock source is used as the reference clock source. However, the disadvantage of this method is that this configuration does not ensure tight synchronization between the transmitter and receiver.



Once the received signal has been downconverted and digitized by the NI PXIe-5622, the LabVIEW program is used to read the digitized I/Q channel data and to display the data in a time domain. A fast Fourier transform is used to view the spectrum of the combined complex data in a frequency domain as well. With the proper setup, periodic physiological signals can be seen in the time domain, and the corresponding peaks can be seen at breathing and heartbeat frequencies in the frequency spectrum.



##IMAGE_2_C##

Figure 1- Block Diagram of the Overall System



Results



Figure 2 shows the measurement results when monitoring a female subject sitting approximately 1 m in front of the radar. The subfigure on the top shows the spectrum of the complex signal at the receiver output, which was obtained using the RFSA Acquire Continuous I/Q VI. A strong peak is seen at the breathing frequency, which indicates the breathing rate of the subject under test was 22 breaths per minute. The second subfigure shows the time domain signal of the radar output for both the I and Q channels. The last subfigure shows the trajectory of the I and Q channels on a constellation graph. With the bioradar system based on LabVIEW and PXI, it is easy to remotely monitor the respiration of a human subject.



Future Plans



Besides the detection of a respiration rate that is demonstrated in this study, accurate measurement of the heartbeat rate and the physiological movement pattern is also important for many health-care and diagnostic applications. To precisely recover the heartbeat rate and original movement pattern, automatic DC calibration is necessary, and more advanced signal processing methods can be implemented in LabVIEW. Taking advantage of the powerful PXI and LabVIEW tools, the smart bioradar system offers unlimited possibilities for signal processing experts, radar engineers, and biomedical engineering professionals to explore further.



Author(s):

Sharmi Banerjee - Texas Tech University

Changzhan Gu - Texas Tech University

Changzhi Li - Texas Tech University