Probing the human body with broadband electromagnetic pulses. Top: Transmitted pulse and received pulse ͑ IRF of the scenario, IRF ͒ which is modulated by the vital functions. Bottom: Reconstructed physiological signature S rec , breathing with superimposed heartbeat, reconstructed from ultrawideband radar data. Anatomic slice taken from Ref. 17. 

Contexts in source publication

Context 1

... transmitter sensor typically an antenna can be used to illuminate the object with microwaves which propagate through the object and are detected by receiver antennas at the other side of the object. Another technique is to look for reflections, using either the transmitter itself or a separate device for detection Fig. 1 radar technique, radio detec- tion, and ranging 1 . The measured data can be processed by suitable reconstruction algorithms to provide information on the complex dielectric permittivity of the scattering object or on the motion of interfaces due to, e.g., human heartbeat and respiration. 2 Noncontact detection and monitoring of human ...

Context 2

... goal of UWB radar is to obtain the impulse response function IRF of a certain object under test Fig. 1. The quality of a measured IRF is mainly determined by the abil- ity to separate closely located peaks and to avoid the mask- ing of smaller peaks due to noise or saturation effects caused by larger signals. The first problem is a question of available frequency bandwidth while the second depends on the dy- namic range of the receiver ...

Context 3

... separate closely located peaks and to avoid the mask- ing of smaller peaks due to noise or saturation effects caused by larger signals. The first problem is a question of available frequency bandwidth while the second depends on the dy- namic range of the receiver and on transmitted power. The classical UWB approach is based on impulse excitation Fig. 1, which implies that the whole transmission chain is subjected to high peak power. Mainly analog circuits tend to overload or saturate in such cases, and system performance degrades. In order to stress the electronics evenly, it is pref- erable to use continuous wideband signals. Typical examples of such signals are swept or stepped ...

Context 4

... some of these requirements can be easily imple- mented in a modified antenna design, the geometrical form and arrangement of conducting and insulating regions re- quire careful consideration, as the radiation properties of the antennas such as radiation pattern or frequency dispersion of the radiated pulses are entirely determined by the excited rf currents. Figure 1 and, in a more abstracted way, Fig. 3 depict the setup commonly used to probe the human body with a UWB device. The body can be assumed to form a multilay- ered dielectric structure with a characteristic reflection coef- ficient f. ...

Context 5

... waves traverse each layer? Furthermore, the influence of physiological processes such as breathing and cardiac motion on the reflection coefficient can be studied. We constructed our model from 14 planar isotropic layers whose arrangement as well as individual thicknesses approximated a transthoracic slice from the vi- sual human data set 17 Fig. 1. The spectral response of a dielectric medium is suitably described in terms of multiple Cole-Cole dispersion Eq. 3 which, with a choice of pa- rameters appropriate to each constituent, can be used to pre- dict the dielectric behavior over the desired frequency range 22 ...

Context 6

... information is also incorporated in the model. For such a layered arrangement Fig. 1, the reflection response, which is equivalent to the transfer function of the object, can be recursively calculated using Ref. ...

Context 7

... this result and further calculations it can be as- sumed that the resolvable motion amplitude lies well below 1 mm. The deviation from the reference can be explained by the compressibility and deflection of the leverage of 3 m length Fig. 11 and by friction-induced sporadic discontinui- ties in the ...

Context 8

... algorithm was tested with stand-alone UWB measurements on a volunteer outside of the MR scan- ner mimicking the envisaged biomedical applications. These tests, particularly aiming to reproduce a breathing paradigm and the superimposed cardiac contraction, exhibit good cor- respondence between calculated time courses and physi- ological reality Fig. 10. The test proved the ability of the algorithm to detect movements in the submillimeter range even for harsh experimental conditions and is therefore well suited for the envisaged ...

Context 9

... demonstrator allows the computer controlled motion of our thorax phantoms inside the MR scanner's head coil operated at 125.3 MHz in the center of the scanner Fig. 11. The motion is driven by a stepper motor over a long leverage 3 m. The nominal spatial resolution was 5.3 m per step. The leverage connecting the motor with the sledge was built of carbon-fiber reinforced plastic to minimize com- pressibility and deflection. As can be seen in Fig. 9, there are deviations between reference profile and ...

Context 10

... of the leverage and by friction-induced sporadic discontinuities in the motion. These uncertainties remain well below 1 mm, however, and are perfectly acceptable. The motional profile of the sand- wich structure was shaped to approximate a respiratory- induced motion of the thorax, superimposed by cardiac os- cillations. This is illustrated in Fig. 12. We utilized the aforementioned M-sequence UWB radar system 4 and proto- types of MR-compatible tapered slot UWB antennas to de- tect the motion of the phantom inside the MR ...

Context 11

... was triggered by the stepper motor controller. Sporadic image artifacts caused by higher order movements acceleration did not hinder the analysis of the MR data which was simply carried out by manual determination of phantom edges in the 128 time frames. Comparing the three independently mea- sured data sets for the position of the phantom in Fig. 12, we find good agreement between the reference profile from the stepper motor controller, the edge positions determined by MRI, and the UWB measurements. Furthermore, the spatial resolution of the UWB radar is similar to that of MRI 1 mm. The correlation between the time sequences de- rived from UWB radar and MRI exceeds ...

Context 12

... the following, the setup for the simultaneous monitor- ing of respiratory and cardiac motion is described. A volunteer was positioned in supine position inside the MR scanner and was breathing spontaneously Fig. 13. The MR-compatible tapered slot UWB antennas T x / R x are po- sitioned about 100 mm above the sternum. A transmit/receive MR surface coil 220 150 mm 2 operating at 125.32 MHz is centered underneath the antennas upon the sternum be- tween the papillae ...

Context 13

... photograph of the MR transmit/receive coil on a body water phantom is depicted in Fig. ...

Context 14

... coil is negligible due to its low dielectric contrast. Reflection signals from the printed circuit board tracks arrive earlier than the body's reflection, and can be separated by our algorithm. The MR transmit/receive coil exhibits a characteristic overall sensitivity profile x , y , z magnitude. For one selected axial slice x , y is depicted in Fig. 14 taken from a MR body water ...

Context 15

... and respiratory motions, we used a one- dimensional imaging technique to speed up data acquisition. This technique produces projection profiles of the integral signal intensity roughly proton density times coil sensitiv- ity along horizontal columns of 2 5 512 mm 3 in an axial slice chosen in an appropriate transversal plane through the heart Fig. 15. This signal reflects the positions of the chest wall and the heart, as the MR signal intensity of the surrounding lung tissue is negligibly small. The resulting profiles are weighted with the sensitivity profile of the MR surface coil which causes the fading of the projection profiles toward the back of the subject Fig. 15, lower ...

Context 16

... through the heart Fig. 15. This signal reflects the positions of the chest wall and the heart, as the MR signal intensity of the surrounding lung tissue is negligibly small. The resulting profiles are weighted with the sensitivity profile of the MR surface coil which causes the fading of the projection profiles toward the back of the subject Fig. 15, lower highly nonlinear motion of the heart during the breath hold. The corresponding one-dimensional 1D projections of the same slice were taken every 50 ms allowing spontaneous breathing, as shown on the lower right of Fig. ...

Context 17

... profile of the MR surface coil which causes the fading of the projection profiles toward the back of the subject Fig. 15, lower highly nonlinear motion of the heart during the breath hold. The corresponding one-dimensional 1D projections of the same slice were taken every 50 ms allowing spontaneous breathing, as shown on the lower right of Fig. ...

Context 18

... temporal evolution of this 1D projection profile is affected by respiration and cardiac motion as shown in Fig. 16 on the left, where the logarithm of the signal intensity is displayed. Cyclic respiratory and cardiac activities are clearly visible along the time ...

Context 19

... we are interested in the motion tracking of se- lected interfaces we observe the variation of R xy over time for each by a covariance analysis Eqs. 11 and 12. The temporal evolution of postprocessed UWB radar signals ver- sus propagation time is given in Fig. 16 right, where the logarithm of the magnitude is displayed. Reflections from the scenario's background static clutter are subtracted from the measurement by extracting the mean value of the ...

Context 20

... applied the same signal processing technique to the MR data set. Figure 17 displays the results of first in vivo experiments applying MRI and UWB radar simultaneously to a volunteer, applying this signal process- ing technique. Cyclic respiratory and cardiac activities are clearly visible along the time course, indicating the correct rendition of the recorded physiological events. ...

Context 21

... this way can uncorre- lated, i.e., linearly independent signals, be analyzed. Our al- gorithm provides this reduction in dimension by an intuitive selection scheme. Another method to reduce dimension by creating uncorrelated signals formed from linear combina- tions of the source signals is carried out by principal compo- nent analysis PCA. 25 In Fig. 18 we present the results given by PCA analysis. For illustrative purpose we used MRI- signatures from different body positions surface y = 100 mm and a deeper position y =80 mm as reference signals to search correlatively in the PCA results calculated from the UWB data set. This was done for the purpose of investigating the motion ...