WO2007141121A1

WO2007141121A1 - New wireless monitoring system

Info

Publication number: WO2007141121A1
Application number: PCT/EP2007/054742
Authority: WO
Inventors: Didier Michel; Raymond Ranwez
Original assignee: Apreco
Priority date: 2006-06-08
Filing date: 2007-05-16
Publication date: 2007-12-13

Abstract

The invention concerns a wireless monitoring system involving the use of passive wireless units working as resonant antenna system with variable resonance frequency (VRS) and a dedicated reader monitoring the variable resonant frequency of the VRS units. The invention also concerns a new type of sensors and a new type of dedicated readers that can be used for such monitoring system (such as for monitoring the various channels for sleep diagnostic and therapeutic analysis).

Description

New wireless monitoring system

Field of the Invention

The present invention relates to the field of monitoring circumstance / situation variables and/or data.

One particular application area of the invention is for instance sleep diagnostic and therapeutic.

More specifically the invention concerns a new type of sensors and a new type of dedicated readers that can be used to control / monitor variables and/or data concerning specific circumstances

/ situations (such as in particular the various channels needed for sleep diagnostic and therapeutic analysis).

Background

A sleep diagnostic process is usually made over a full night and concerns the recording and the interpretation of many different types of channels such as: respiratory movement, air flow, body position, eye movements, ECG, EEG, EMG, oxygen saturation, leg activity, respiratory noise, etc.

A classical sleep study requires 20 to 40 sensors that are placed on the patient's body and connected to various devices through wires. These processing units carry out the amplification, filtering, analog to digital conversion. The storage, analysis and display are usually performed by a computer.

A physician working in a sleep laboratory is always concerned with the patient's comfort in the hospital in order to get a real representation of the sleep at home. The number of wires needed to connect each sensor to the front-end device severely constrains patient's mobility in the bed and could be viewed as a strong bias to the sleep study. This whole sheaf of wires can also induce stress on the sensors and could imply channel disconnection or connector damage.

Today different wireless technologies are available on the market and many adaptations using these technologies are made fighting for the "Patient Liberation". Currently the wireless signal is used for data transmission between a small front-end collecting the information and performing the digitization process and a central recording device. The patient is in this case free to move around by carrying a "little box" but the number of wires attached to the sensors and connected to the front-end is still the same.

Some systems have been described, where the signal is digitized on the sensor itself, then sent with some kind of a digital transmission (WiFi, RFID, ...). These systems imply a power source at the sensor location, in order to amplify and/or filter and/or digitize the original signal.

The power source on the sensor means a lot of complications, because of the short lifecycle or the need to recharge. Description of the invention The present invention may be seen as an analog network system transferring the information wirelessly and without mandatory local power supply at the sensor location.

The system involves the use of passive wireless units (at least one passive wireless unit working as resonant antenna system with variable resonance frequency, herein referred to as "variable resonance sensors" VRS), sensors (at least one) measuring circumstance / situation variables and/or data, capable to act upon said VRS unit(s), and a dedicated reader monitoring the variable resonant frequency of said VRS unit(s). The analog treatment and digital process can be made on the reader side.

In particular the invention can involve sensors each related to a physiological activity and a dedicated reader monitoring the variable resonant frequency of several VRS units. DETAILED DESCRIPTION

The basic idea is to make a resonant antenna system on the sensor side, for which the resonant frequency will vary according to the signal to be measured. One RLC circuit (called "reader") is excited by an alternative source of tension at a frequency close to its resonant frequency. The induced electromagnetic field oscillates at the same frequency.

Another RLC circuit (called "sensor") having a resonant frequency close to this of the reader is placed in the electromagnetic field. If the sensor's inductance (the antenna) is correctly positioned (orientation, distance,...), an efficient inductive coupling appears between the 2 circuits. This effect can be monitored in the reader, for instance by measuring the tension at the terminal of the inductance or by measuring the current.

If the physiological parameter to be observed can act on the sensor in such a way that the electrical characteristics of the sensor are modified and/or the mutual inductance between the 2 circuits is modified, then the parameter will be observable on the reader side, wirelessly. In particular, no battery is needed to power the "wireless transmission" from the sensor to the reader. The reader or primary circuit (see Fig. 1), is composed of a resistance ^p, a capacitor C_p and an inductance ^p. The circuit is excited by an alternative sinusoidal tension generator at an angular frequency ω. The value of the tension over time is:

U_p(t) = U₀ cos ωt = U_Qe^3wt Z₁ \ The sensor or secondary circuit (see Fig.1 ) is composed of a resistance Rs, a capacitor C_s and an inductance L_Sm There is no direct source of tension in the circuit. The two circuit are inductively coupled, with a mutual inductance M. Note that

AI = k ^/ L_pL_s, _QΛ k being the coupling coefficient.

According to Kirchhoff's laws, the currents Ip and ^s obey the following system of equations:

The source of tension being sinusoidal, the equations above can be expressed as

^Rvh ^~ J-^r¹ _P + JωLpIp - jωMI_s = U_p

R.J_h - j —— I, + jωLJ_h - jωMI_p - 0

UJ Cs (₆)

Let us define the primary and secondary impedances as

^ωϋ» / (8) The equations of the circuits are now reduced to

Z_pI_p — jωMI_s = U_p ^

Z₈I₈ - jωMI_p = 0 ^ _j Q)

Because we will be making our measurements on the reader, we are interested in the primary:

j _ U_p

P ₇ , ω²M*

P ^ Z₆ (H)

If the sensor is absent (or far away from the reader), then M — 0 and Eq. (1 1 ) simplifies into this of a simple RLC circuit:

T _ Uv ^ZP (12)

The absolute value of the primary current is the modulus of (13):

U_n

¥P \\ =

R_p ² + (ωL_p - 1

;C_V

(14)

Figure 2 shows a plot of

is maximum when the denominator of (14) is

. T 1_ _ A minimum, i.e. when ¹^ P ^C₁, ^J. This occurs for an angular excitation frequency ^₀, called resonant frequency:

1

L_pC_p (₁₅)

The bandwidth Aω of the circuit is defined as the range of frequencies where the current is above ^~/T^~ (-3dB). The quality factor Q is the ratio of the resonant frequency to the bandwidth:

UQ

Q = Δω (17)

^RP (19)

Let's now recall our complete system (reader + sensor). The characteristics L_p and C_p of the reader are chosen so that the resonant frequency ^ωo_P of the reader matches the frequency of the generator ω\ the reader is tuned. The characteristics L_s and C_s of the sensor are chosen the same way: the sensor is also tuned.

⁼ ^ ⁼ V€ (2!) The primary and secondary impedance from equations (7) and (8) become:

Z_n — R_r (22)

The equation (11 ) for the current in the reader simplifies into :

/ -¹ PtUIiPd = 1¹PmIn = _p ^U» ω'² M'² ^ΆP + ^~1ΪΓ^~ (24)

We clearly see that the presence of the tuned sensor makes the current in the reader drop. In fact

is the absolute minimum of the current (for a given mutual inductance M). Hence, detuning the sensor will lead to a bigger Z₃ (see Eq. (8)) and increase the current (see influence of Z_s jn Eq. (11 )). Sensor detuning will be investigated further in the next section. One can also observe that a tuned sensor does not induce a dephasing between tension and current in a tuned reader (the imaginary part of (24) is null).

It is now possible to study the variation of the current in the reader as a function of the coupling coefficient k. Using Eq. (2) and (17), we can rewrite Eq. (24) as:

^p- R_p (1 + k.ⁱQ_pQ_s) ₍₂₅₎

Equation (25) demonstrates that to maximize the influence of the sensor on the reader's current, we can: • maximize the coupling coefficient (Figures 3 - 4)

• maximize the quality factors The coupling coefficient can be evaluated using a commonly used formula:

with ^rp the reader's antenna radius, i"s the sensor's antenna radius, d the distance between antennas, θ the angle between antennas. Figure 5 shows the coupling coefficient of a reader antenna of radius 25cm and a sensor antenna of radius [1..5cm]. Combining (25) and (26), we can plot (Fig. 6 t,_π as a function d, keeping the same parameters as for Fig. 4 and Fig. 5.

As the distance increases, k ~ 0 and the current in the reader approaches this of a simple RLC circuit (see previous section) :

_ If we let the physiological parameter Φ change a characteristic in the sensor circuit, say the capacitor C₃₁ the sensor will lose its tuning: the sensor is not anymore excited by an alternative magnetic field oscillating at the right frequency (i.e. the resonant frequency of the sensor). The reader, however, stays tuned. Equation (11) becomes :

ⁿP ⁺ z_s (27)

Comparing to this of a tuned sensor, the impedance Z_h{ώ) of the sensor could not be simplified into its resistive part: the reactance of

^p °* \ ^{s ωfϊ}«W/ is not null because of the detuning. This has 2 consequences:

- The detuning implies that

and

^JPtun_Hd < ll^Jp(» ll ≤ ⁷Pn_13x (29)

- The physiological parameter and its variations are thus observable on the reader side by measuring the amplitude of the current (or any related value).

Since the imaginary part of the current equation is not null, a tension-current dephasing appears. This effect is shown on Fig. 8. While we can think of measuring this dephasing to observe the physiological parameters, there are 2 obstacles :

• The amplitude of the dephasing is so small that it could be difficult to measure.

• Near ^tuned - where the amplitude is maximum - the dephasing curve is not monotonic, even if C_h(φ) \_s kept at one side of C*_tuned

( Cpiø) < C,. _{JlW or} CJΦ) > C._Vιinoα)_ jhjs means that we may not be able to distinguish between two distinct states of the physiological parameter.

The above leads to possible implementations in respect of :

- variation of the coupling factor

Reader and sensor are tuned at the same frequency. The sensor is placed on an object or body. Movements of the object or body affects the coupling coefficient k by modifying the distance reader- sensor or the relative orientation of the 2 antennas. The variations of k engender variations of the current in the reader. These variations (or any other related value) are measured. • The mean distance between the reader and the sensor should be chosen where the sensibility is maximum (small movements will produce great signal variation) and signal measurable. In the example presented in Fig. 6 this distance is about 0.5m.

• Only movements of reasonable amplitude ([1cm..50cm]) will be measurable. • Only usable for measuring movement and/or change of orientation. Impossible to distinguish between movement and changes of orientations

• Array of sensors possible if the sensors are tuned at sufficiently different (bandwidth..) frequencies. Then multiple readers (each being tuned with a sensor) are used or a single tunable reader interrogating one sensor at a time is used.

- Oscillator tuning/detuning the sensor at a variable frequency

Reader tuned, sensor is tuned/detuned at a variable frequency, the frequency being a function of the physiological parameter.

On the sensor side, the physiological parameter is actually frequency modulating an oscillator which in turn acts on the sensor's basic RLC circuit by short-circuiting (sensor highly detuned) or not (sensor tuned) the capacitor (see Fig 9). On the reader side, the frequency modulated signal is demodulated to extract the physiological signal information.

Figure 10 shows an example of electrical schematic for the sensor. Here, the physiological parameter acts on a varicap. The change of value of the capacitor changes to oscillating frequency of the oscillator. The oscillator needs a power source. Since the power required is very low, one can consider several options for the power source:

• Typical 3V lithium battery. Would last for years (5).

• Super Cap. Would be charged via the sensor inductor by putting the sensor near an appropriate inductive charger when the sensor is not used. Would be of enough capacity to power the sensor for several hours. • Direct use of the power generated by the reader in the sensor's inductance (cf. RFID).

Remarks : + For a given reader-sensor distance, the maximum SNR is always achieved. + Easily allows for multiple sensors if each sensor has a different carrier wave base frequency (each sensor has its own "channel"). + Possibility to extract distance/movement information from the amplitude of the FM signal. Also possible to fire "bad reception" warnings if the amplitude gets too low.

- Sensor detuning as a function of the physiological parameter

Reader tuned, sensor basically tuned at the same frequency. The physiological parameter affects a component of the RLC circuit, for instance the capacitor. The sensor is thus slightly detuned, the amount of detuning varying with the physiological parameter. There are several ways to measure the sensor detuning on the reader : Measuring the amplitude of the current The sensor detuning affects the current in the reader and any other related value (tension,...). Theses values can be measured.

• As the current is also a function of the distance reader-sensor, variations of this distance can interfere with the signal of interest (see section on variation of coupling factor) • An array of sensors is possible as explained in the section on variation of coupling factor. Measuring the dephasing between tension and current

As explained in section "reader tuned, sensor varying", a detuned sensor induces a dephasing between the tension and the current in the reader. • As dephasing is independent of the distance reader-sensor, variations of this distance doesn't interfere with the signal of interest

• An array of sensors is possible as explained in section on variation of coupling factor)

Searching the tuning frequency of the sensor By frequency sweep :

The generator in the reader sweeps a range of frequencies to find the resonant frequency of the sensor. The resonant frequency of the sensor is the frequency for which the current in the reader is minimum.

• The reader design is more complex. In particular, the reader must stay tuned during the frequency sweep of the generator to achieve a good sensitivity. The reader's RLC circuit characteristics must thus be changed dynamically according to the frequency of the generator.

• The number of frequency sweeps (or sweep cycles) per second determines the sampling frequency of the signal of interest. • The signal of interest is less affected by the movements of the sensor

• An array of sensors is possible if the sensors operates in separate frequency bands. The reader sweeps in a band, then hops to another band, resumes its sweeping, aso.

By listening to the sensor's damped resonant frequency The reader charges to sensor's RLC circuit by emitting an oscillating magnetic field close to the sensor's presumed resonant frequency. The generator is then switched off. The sensor will continue to oscillate for some time (depending on the damping factor) at it's own damped resonant frequency. The reader listen to the oscillating magnetic field emitted by the sensor and monitor the frequency.

• The damping factor is *^■> ⁼ 2T

• The damped resonant frequency is slightly different from the natural resonant frequency: C²

Possible VRS implementations for the different physiological functions that can be obtained from a patient are provided in the following table. Note that a VRS (Variable Resonant Sensor) comprises a VRU (Variable Resonant Unit) and a sensor.

TABLE 1 Possible VRS implementations for different physiological functions

The following comments can be made on the above stated sensor methods

* Actimetry The actimetry function could be either analog or digital. In the digital way, the sensor is sensitive to an amount of acceleration and triggers when this acceleration is above a certain threshold. In the analog way, it means that the analog data could be proportional to a movement or to the acceleration of a part of the body. Ball-based actimetry, digital (cf. fig.11 )

A VRS is placed on a leg. The resonant frequency of the VRU is changed just by switching the VRU's antenna, or by switching a capacitor placed on it by a "classic" ball-based actimeter.

The movement can simply be detected by analyzing the amplitude modulation on the reader antennas when a movement occurs. Piezo-based actimetry, analog (cf. fig. 12)

A VRS is placed on a leg. The resonant frequency of the VRU is modified by means of a Varicap itself biased by the piezo sensor.

The relation between the Varicap capacitance and the applied voltage is given in figure 13. * Body Position

The body position function could be based on the same principle as for the actimetry function. Ball-based body position function A VRS is placed on the patient body (for example on the chest). The VRU's frequency is changed just by switching the VRU's antenna, or by switching capacitors placed on it by a "classic" ball-based body position sensor. The VRU has one frequency per position. Gravity-based body position A VRS is placed on the patient body (for example on the chest). One or more tilt switches change the VRU's frequency.

* Respiration movements

Piezo-based VRU (cf. fig. 14)

A piezo belt drives a Varicap which in turns changes the VRU's resonant frequency

Using variable capacitor (cf. fig. 15)

The sensor part of the VRS is a capacitor formed inserting a sheet of copper between 2 sheets of foam and an external shield. Using variable inductor (cf. fig. 16) An inductor (a coil) is placed around the patient like a belt so that its inductance changes with the respiration movement. This, in turn, changes the VRU's resonant frequency (cf. figure 16).

* Respiration, airflow

The sensors described here are based on a thermistor, a pressure cannula and on a new mechanical design. Thermistor-based VRS

A possible implementation is to bias a Varicap, which controls the VRU resonant frequency, by the voltage drop created through the thermistor, the voltage divider being powered with a battery. Other schemes are possible.

Pressure sensor-based VRS

The pressure sensor can be either a strain gauge or a piezo. They can be placed in the flow (like a Venturi or a device comprising a fine- mesh screen in a tube) or orthogonally to the flow (like a Pitot tube). Both need at least amplification and a battery. The output could then drive a VRU's Varicap. Special mechanical design. The pressure created by the airflow may change the distance between two metallic flaps forming a capacitor which in turn changes the VRU's frequency.

* Sounds

The sensors described here detect sounds (like snoring) from the patient.

Piezo snore microphone

A piezo microphone (or other kind of microphone) is used to get the snoring noise from the patient. This sensor (probably) needs amplification before acting upon a VRU's Varicap. Capacitive snore microphone

The sound coming from the patient acts on a metallic plate that vibrates with the sound. This plate could be part of a capacitor that changes the VRU's frequency.

* Cardiac sounds The same kind of sensors as described above for snoring detection could be used to get cardiac sounds. * Electromyography (EMG) EMG using surface electrodes

Normally, EMG needs differential amplifier. Since the VRS is inherently isolated from any other device, and since the EMG channels do not need to be interconnected (one channel per VRS), a simple amplifier (2 or 3 stages) could be enough to drive a VRU's Varicap and give a valuable signal. This needs a battery.

^* Cardiac functions

ECG A one-channel ECG could be done like a one-channel EMG

VRS. If more than one channel is needed then the signal should be multiplexed. Circulatory plethismography

This kind of VRS could use a chamber in which the pressure is measured with either a variable capacitor formed by moving metallic plates, or a variable inductor. This capacitor or the inductor can then change the VRU's frequency. Oximetry

There are two options for the oximetry VRS. The first is to calculate the SpO2 value within the VRS with a microcontroller and to send the calculated SpO2 value (in addition to the plethismographic data) digitally through the VRU. The second option is to send the raw data (Red base line, Red delta, IR base line, and IR delta) in a multiplexed fashion through the VRU, the SpO2 value being calculated on the reader side. In both cases, the gain of the VRS is the reduced power needed internally since the communication do not need any power. * EEG

Using surface electrodes

Like for oximetry, the gain brought by the VRS technology is on the battery side since there is no power needed for the wireless communication. Except for that point, the EEG part of an EEG VRS is classic. * EOG

Using surface electrodes

This is very similar to an EMG channel if a single differential channel is needed (one electrode per eye). Using a piezo sensor

In terms of the electronics being used, this VRS is very similar to a snore microphone except for the bandpass (longer time constant).

Claims

1. Wireless monitoring system for circumstance / situation variables and/or data, characterized in that the system involves the use of at least one passive wireless unit working as resonant antenna system with variable resonance frequency (VRS), at least one sensor measuring circumstance / situation variables and/or data, capable to act upon said VRS unit(s), and a dedicated reader monitoring the variable resonant frequency of said VRS unit(s).

2. Wireless monitoring system according to claim 1 , characterized in that the system involves at least one sensor measuring a physiological activity.

3. Wireless monitoring system according to any one of claims 1 and 2, characterized in that said VRS unit involves a receiver emitting a fixed frequency and monitoring the variations of the absorption by the sensor when its oscillation frequency moves away form the emitter's one.

4. Wireless monitoring system according to any one of claims 1 and 2, characterized in that said VRS unit involves a receiver emitting a sweeping frequency, around the pre-determined sensors oscillation frequency, and measuring the absorption of the signal at all frequencies, to find the maximum absorption frequency.

5. Wireless monitoring system according to any one of claims 1 and 2, characterized in that it involves a sensor comprising an oscillator with a frequency modulated in function of the measured circumstance / situation variables and/or data, and acting on the resonance frequency of said resonant antenna system, and a reader able to demodulate said modulated frequency to extract said measured circumstance / situation variables and/or data.

6. Passive wireless unit working as resonant antenna system, characterized in that the resonant antenna system involves a variable frequency.

7. Passive wireless unit according to claim 6, characterized in that a RLC circuit capacitance of the antenna system or an oscillator circuit frequency acting on the resonance frequency of the antenna system can be changed by:

- two metallic parts (forming a capacitor) moving so that the surface in common changes with the parameter, or - two metallic parts (forming a capacitor) moving so that the distance between them changes with the parameter (as by compression), or

- changing the characteristic of dielectric insulation between the plates.

8. Passive wireless unit according to claim 6, characterized in that a RLC circuit inductance of the antenna system or an oscillator circuit frequency acting on the resonance frequency of the antenna system can be changed by

- changing the size or length of the coil, or

- changing the reluctance of the magnetic circuit.

9. Passive wireless unit according to claim 6, characterized in that a quality factor of the antenna system (or of one of its components) can be changed by changing a parameter of said antenna system, e.g. by placing a variable resistor (such as a thermistor) in parallel or in series with an oscillator circuit of the antenna system or a RLC circuit of the antenna system.

10. Passive wireless unit according to claim 6, characterized in that a quality factor of the antenna (or of one of its components) can be changed indirectly by a Varicap, which in turn is driven by a sensor that gives a voltage, such as a piezo-electric belt.

11. Passive wireless unit according to any one of claims 6 to 10, characterized in that it is used in a wireless monitoring system according to any one of claims 1 to 5.

12. Reader for a resonant antenna system, characterized in that it involves a fixed resonant frequency and that it is sensitive to the absorption of its emitted electromagnetic field by a passive wireless unit according to any one of claims 6 to 11.

13. Reader for a resonant antenna system, characterized in that it involves a variable resonant frequency, which "scans" the frequency range absorbed by at least one passive wireless unit working as resonant antenna system.

14. Reader for a resonant antenna system, characterized in that it involves "frequency hopping" where the resonant frequency changes by steps.

15. Reader for a resonant antenna system, characterized in that it involves "tracking" the frequency of a passive wireless unit working as resonant antenna system and locking on said frequency so as to follow any changes in the resonant frequency.

16. Reader according to any one of claims 13 to 15, characterized in that the resonant antenna system is comprised by a passive wireless unit according to any one of claims 6 to 11.