A cerebral edema monitoring system based on a new excitation source

Abstract

BACKGROUND:

Real-time clinical monitoring of cerebral edema (CE) is of great importance and requires continuously improved and optimized measurement hardware.

METHODS:

A new excitation source with higher frequency stability and wide output power range is presented in this work. The proposed excitation source is small in size and easy to integrate. The output power range of excitation signal used is 1.5 $\sim$ 33 dBm with a reference signal of 9 $\sim$ 11 dBm, and the phase shift stability of the excitation signal and reference signal reach 10 ${}^{-7}$ within 20 min.

RESULTS:

When normal saline (0.9%, 10 mL, 20 mL, 30 mL, 40 mL, and 50 mL) is injected into a human head phantom model, the magnetic induction phase shift (MIPS) changes from 252.78 $\pm$ 7.61 degrees to 252.40 $\pm$ 7.77 degrees. The MIPS signal shows a downward trend with increasing volume, indicating that MIPS can reflect the volume change of the measured object. Moreover, a more dramatic trend is visible when the solution volume increases from 0 to 10 mL and from 40 to 50 mL. This occurs where the volume increment is closer to the upper and lower sides of the over-ear sensor, where the magnetic field is strongest.

CONCLUSION:

The phantom simulation experiments illustrate that the proposed MIPS detection system based on a signal source can detect the real-time progress of CE. Advantages of low cost, high precision, and high sensitivity endow this system with excellent application prospects.

Keywords

Excitation source cerebral edema phase shift magnetic induction

1. Introduction

Stroke mortality in China accounts for one-third of stroke deaths worldwide [1]. Cerebral edema (CE) is a secondary brain injury caused by the occurrence and development of a stroke which can result in increased intracranial pressure (ICP), brain line displacement, brain hernia, and even death, due to increased water content in the brain tissue [2]. Clinical CE monitoring is therefore of great significance. Currently, CE monitoring mainly relies on imaging devices such as computed tomography (CT) and magnetic resonance imaging (MRI), which provide one-off evaluation rather than bedside real-time monitoring [3, 4, 5]. A number of new methods for noninvasive detection of brain edema including transcranial doppler ultrasonography (TCD), flash visual evoked potential (FVEP), near infrared spectroscopy (NIRS), optic nerve sheath diameter (ONSD), and electrical impedance tomography (EIT) have emerged in recent years [6, 7, 8, 9, 10]. However, these novel methods also contain a number of issues that remain unresolved. For example, TCD consistently shows a poor correlation when ICP reaches a high level and also requires professional operation skills, while FVEP is not sensitive in the early stage of acute CE where ICP shows slight change. Additionally, EIT requires direct contact between the electrode and skin to inject current and it is difficult for the current to pass through the skull due to its high resistivity, which seriously affects the imaging quality [11, 12]. Thus, the available noninvasive methods cannot provide accurate real-time and bedside monitoring.

Magnetic induction phase shift method (MIPS) has been studied as a potential tool for CE noninvasive monitoring in recent years by scholars including Hart, Griffiths, Scharfetter, Manoufali, and Gonzalez [13, 14, 15, 16, 17]. Using this method, a certain frequency range of magnetic fields is applied to induce eddy current in cerebral tissues. The phase shift ( $\Delta\theta$ ) spectroscopy related to cerebral tissue information between the excitation magnetic field and the inductive magnetic field is then detected. The conductivity of biological tissues is notably low ( $\sigma<$ 3 Sm ${}^{-1}$ ) and the induced magnetic field is typically only 1% of the primary magnetic field at a frequency of 10 MHz [18]. Highly accurate MIPS measurement hardware is thus required and our previous works were devoted to designing an MIPS detection system with high sensitivity and high stability. In earlier studies, an excitation source for a MIPS detection system was developed with frequency stability of less than 10 ${}^{-4}$ in some frequencies and output powers [19, 20]. In 2014, we established a magnetic induction phase detection system which could achieve phase noise as low as 6 m ${}^{0}$ and a 4 h phase drift as low as 30 m ${}^{0}$ at 21.4 MHz [21]. The MIPS of cerebral hemorrhage in rabbits with a single coil-coil was also investigated [22]. A broadband phase shift detection system was subsequently designed and completed by WC Pan in 2015 [23], while Li designed a set of MIPS brain edema real-time continuous monitoring systems with high sensitivity and good stability in 2017 [24]. Based on these continuously improved and optimized systems, our research team has made some progress in monitoring CE and cerebral hemorrhage in animal experiments [25, 26].

Continuously improved and optimized measurement hardware is necessary for CE real-time monitoring. In this work, an excitation source with higher frequency stability and wide output power range is designed. The signal source, an over-ear sensor, and a PCI acquisition card are combined with LabVIEW to build a real-time CE detection system. The performance of the system is then tested and assessed using a conductivity resolution experiment and phantom simulation experiments.

2. Methods

2.1 Design of the fixed frequency detection system

2.1.1 Signal source

As illustrated in Fig. 1, the excitation source contains six components: a direct digital synthesizer (DDS) signal generator, frequency doubling and amplifier circuits, an excitation signal and reference signal generator, a power amplifier circuit, a control interface circuit, and the input interface and display circuits.

Figure 1.

Working principle diagram of excitation source.

An excitation source requires high frequency stability to perform accurately in CE detection systems. In addition, the output power of both excitation signal and reference signals must be easily adjusted. When the frequency and output power of the excitation signal and reference signal is set up, their phase shift should ideally remain invariant over a long period of time (hours). The proposed excitation source combines two distinct technologies, as illustrated Fig. 2. The performance and feature characteristics are composed of seven 1.5–33 dBm sinusoidal signals of excitation signal, seven operation frequencies with reference signal of 12 dBm, a frequency and output power selection and display panel, BNC interface, USB for PC software connection, a remote control, and a 220 V power supply.

Figure 2.

Physical diagram of excitation source.

2.1.2 PCI acquisition card and LabVIEW software platform

The PCI acquisition system (National Instruments Corporation, USA) is widely used in various types of data acquisition applications. In this work, the sampling rate of the PCI-5124 acquisition card was set to 100 MHz using the software platform LabVIEW2014 (National Instruments Corporation, USA). The number of sampling points was set as 400,000, and the data obtained by the acquisition card was used to display the results of software phase detection. Additionally, the fast Fourier transform (FFT) method was used for phase detection.

2.1.3 Sensor

The over-ear sensor is comprised of an excitation coil and a detection coil. If necessary, any coil can be used as an excitation coil, while the other coil can be used as a detection coil. As shown in Fig. 3, the sensor was modified from ordinary headphones, where parts of the earphones were replaced by coils. Both coils were copper-painted covered wires (wire diameter 0.8 mm) wound 10 turns, and were closely arranged and well insulated. The diameter of the two coils was 8 cm, and the space of the coaxial placement was 20 cm. This wearable sensor is comfortable for long-time monitoring.

Figure 3.

The over-ear sensor.

2.2 Performance test of the system

2.2.1 Signal source test

The temperature was set at 20 ${{}^{\circ}}$ C and all tests were carried out after the MIPS system was turned for 20 min. The frequency stability of the excitation signal was measured at 1.5 dBm, 7 dBm, 12 dBm, 15 dBm, and 24 dBm frequency points, while the reference signal was measured at 12 dBm. The reference signal voltage was less than the detection threshold of $V_{pp}=$ 5 V according to a Tektronic FAC3003 frequency meter with accuracy of 10 ${}^{-9}$ . Every frequency was measured 1000 times for 20 min, longer than the single test time of CE detection systems based on MIPS [24], the $f_{\max}$ and $f_{\min}$ was determined, then frequency stability was processed as follows:

$\displaystyle f_{\Delta}=\frac{f_{\max}-f_{\min}}{f_{0}}$

As illustrated in Fig. 4, the phase shift measurement was also successfully performed by using NI-PXI5124 and LabVIEW 2012. The correlation method employed for the phase discrimination algorithm has the advantage of high precision and fast speed. Every frequency was measured three times for 20 min. To obtain an optimal combination of frequency and output power for the CE detection system based on MIPS, a variance analysis of non-repeated dual factors test was used. The variance analysis principle is as follows: there are two variables A and B in a test. A has seven levels which are A ${}_{1}$ , A ${}_{2}$ , …, A ${}_{7}$ (A ${}_{1}=$ 200 kHz, A ${}_{2}=$ 1 MHz, A ${}_{3}=$ 10.7 MHz, A ${}_{4}=$ 21.4 MHz, A ${}_{5}=$ 30.85 MHz, A ${}_{6}=$ 40.05 MHz, A ${}_{7}=$ 49.95 MHz), while B has four levels which are B ${}_{1}$ , B ${}_{2}$ , …, B ${}_{4}$ (B ${}_{1}=$ 10 dBm, B ${}_{2}=$ 16 dBm, B ${}_{3}=$ 22 dBm, B ${}_{4}=$ 28 dBm), respectively. If one test is carried out according to each kind of combination level (A $i$ , B $j$ ), and the result is $x i j$ ( $i=$ 1, 2, …, 7; $j=$ 1, 2, …, 4), then all the $x i j s$ are independent of each other and the normal distribution should be followed. For each of the test values $x i j$ , $i$ represents the corresponding level of factor A, while $j$ represents the corresponding level of factor B. The total number of the test results is $n=rs$ . The reciprocal of slope (1/K) reflects the stability of signals under different frequencies and output powers, so it is repeated three times and the mean value is determined as the end value. The greater the number of 1/K, the better the signals. In this work, the number of 1/K is 10 ${}^{7}$ and $k$ is the same stability degree as the excitation and reference signal, so in the process of datum processing, all datum are divided by 10 ${}^{7}$ .

Figure 4.

The measurement system.

2.2.2 Phantom simulation experiment design

The excitation frequency of the new excitation source was set to 49.95 MHz and the output power was 1.78 W (32.5 dBm). Normal saline (0.9%, 10 mL, 20 mL, 30 mL, 40 mL, and 50 mL) was injected into one human head phantom model. Each liquid was measured 16 times, and the average data value was obtained. The excitation signal was connected to the right earphone port of the skull model, the left earphone port coupling signal was the input to the PXI acquisition card, and the phase difference between the reference signal and the signal coupled to the left earphone port was measured by the PXI acquisition card phase detector. This phase difference can reflect changes in the brain. A beaker was placed in the middle of the skull model, and 10 mL, 20 mL, 30 mL, 40 mL, and 50 mL volumes of physiological saline were injected to simulate cerebral hemorrhage under different severity conditions. The phase difference data under different conditions was then separated. Each measurement was 10 s long, and the average value was taken as a single measurement phase difference data. This average value is the final value of the phase difference corresponding to each cerebral hemorrhage condition. A plot of the injected volume and phase difference was then drawn.

2.3 Statistical analysis

All data was expressed as mean $\pm$ standard deviation from at least sixteen independent experiments. The exponential decay function was then selected to carry out the non-linear regression analysis in Origin 9.1 (Origin Lab, Massachusetts, MA, USA). The values of all regression coefficients were determined, and the variance significance F test was carried out. The significance level was set at $p<$ 0.05.

3. Results

3.1 Performance test of the signal source

The frequency stability of excitation signal test datum and the variance of excitation signal test datum are provided in Tables 1 and 2, respectively. Figures 5 and 6 show excitation signal frequency stability and frequency stability variance at 1.5 dBm, 7 dBm, 12 dBm, 15 dBm, and 24 dBm frequency points. According to the curves, frequency stability is at 10 ${}^{-7}$ level. The frequency stability is in nearly the same level at the same frequency point, except for the 200 kHz point. In the same output power, excitation signal frequency stability variance is higher when frequency changes from 200 kHz to 49.95 MHz.

Table 1
Frequency stability of excitation signal test datum

Frequency	Frequency stability (10 ${}^{-7}$ ) of excitation signal
	1.5 dBm	7 dBm	12 dBm	18 dBm	24 dBm
200k	1.51	1.23	0.96	0.94	0.55
1M	0.55	0.33	0.39	0.23	0.28
10.7M	0.28	0.25	0.32	0.32	0.29
21.4M	0.26	0.24	0.33	0.34	0.26
30.85M	0.34	0.27	0.31	0.34	0.29
40.05M	0.29	0.27	0.31	0.29	0.25
49.95M	0.29	0.25	0.29	0.29	0.24

Table 2

Variance of excitation signal test datum

Frequency	Variance (mHz) of excitation signal
	1.5 dBm	7 dBm	12 dBm	18 dBm	24 dBm
200k	6.01	3.45	2.25	2.67	1.61
1M	7.65	5.87	8.38	6.13	5.75
10.7M	61	60	72	74	63
21.4M	114	113	147	124	117
30.85M	207	163	212	190	169
40.05M	228	212	235	243	214
49.95M	281	261	292	299	276

Figure 5.

Excitation signal frequency stability.

Figure 6.

Excitation signal frequency stability variance analysis.

3.2 Phantom simulation experiment

Normal saline (0.9%, 10 mL, 20 mL, 30 mL, 40 mL, 50 mL) was injected into one human head phantom model and the MIPS changed from 252.78 $\pm$ 7.61 degrees to 252.40 $\pm$ 7.77 degrees. The change of MIPS during the 0–60 mL saline injection process is illustrated in Fig. 7. When the saline volume change from 0–50 mL, MIPS shows a downward trend with a mean variation range of 0.382 $\pm$ 0.272 ${}^{\circ}$ ( $n=$ 16). It can be clearly observed that the MIPS signal shows a downward trend with increasing volume, indicating that MIPS can reflect the volume change of the measured object. This result was similar to findings in our previous work, indicating that the decrease of MIPS is caused by the increase of volume [22, 23, 24].

Table 3 shows the relationship of MIPS and volume change. When the first 10 mL saline is injected into the head model, MIPS changes from 0.1367 degrees. Meanwhile, MIPS changes to 0.0212, 0.0265, 0.0686, and 0.1290 degrees over the next four injections. A more dramatic trend is also observed when the solution volume increases from 0 to 10 mL and from 40 to 50 mL. This occurs in cases when the volume increment is closer to the upper and lower sides of the over-ear sensor, where the magnetic field is strongest.

Table 3
Relationship of MIPS and volume change

	Volume change (mL)	MIPS change (degrees)
1	0–10	0.1367
2	10–20	0.0212
3	20–30	0.0265
4	30–40	0.0686
5	40–50	0.1290

Figure 7.

MIPS change with the different injected saline volume.

4. Discussion

The clinical real-time monitoring of CE is of great significance, with hardware measurement systems requiring constant improvement and optimization to carry out this non-invasive measurement. An excitation source with higher frequency stability and wide output power range was designed in this work. The measurement system was mainly comprised of an excitation source, sensor, PCI acquisition card, and the LabVIEW platform. The performance of the proposed system was experimentally assessed using a conductivity resolution experiment and phantom simulation experiments.

The results of the performance study indicate that the new excitation provided a wide output power range, good frequency stability, and a high phase shift stability. The output power range of excitation signal was 1.5 $\sim$ 33 dBm and reference signal was 9 $\sim$ 11 dBm. The phase shift stability of the excitation signal and reference signal was found to be 10 ${}^{-7}$ within 20 min. Compared with our previous measurement system, the new excitation source can output optimal frequency and power for different measurement objects, so that the measurement system achieves optimal performance [25, 26].

The over-ear sensor was constructed using an excitation coil and a detection coil, and is comfor- table for long-time monitoring. To verify the feasibility of the proposed new system for brain edema measurements, normal saline (0.9%, 10 mL, 20 mL, 30 mL, 40 mL, and 50 mL) was injected into one human head phantom model. Each liquid was measured 16 times, and average data value was obtained. The MIPS signal showed a downward trend with increasing volume, indicating that MIPS can reflect the volume change of the measured object. This result was similar to our previous work, indicating the decrease of MIPS is caused by the increase of volume [22, 23, 24].

A more dramatic trend was observed when the solution volume increased from 0 to 10 mL and from 40 to 50 mL. This occurred where the volume increment was closer to the upper and lower sides of the over-ear sensor, where the magnetic field is strongest. The results of this experiment are also consistent with our previous findings, suggesting that MIPS measurements are more sensitive to volume changes in physiological measurements. Therefore, in contrast to imaging methods such as CT or MRI, it can be concluded that MIPS method can detect small volume changes in the brain and provide early detection of small lesions in the brain. This method can help patients obtain early diagnosis and treatment, as well as reduce the death rate and disability rate of CE.

The new excitation source is small in size, high in frequency stability, and easy to integrate. Simulation experiments illustrate the reliability and feasibility of the detection system. However, further improvements are required in the future such as the inclusion of a temperature control module to avoid common mode interference. Subsequent work will focus on improving the system to increase its frequency band to cover the millimeter wave and microwave bands, and improve its application in bioelectromagnetic measurement. Animal experiments and volunteer experiments are also required to further verify the reliability of the system.

5. Conclusion

An excitation source with higher frequency stability and wide output power range was designed in this work. Phantom simulation experiments demonstrated that the proposed MIPS detection system based on a signal source can detect the real-time progress of CE. Advantages of low cost, high precision, and high sensitivity endow this system with significant application prospects.

Footnotes

Acknowledgments

This work was financially supported by the Equipment Project (No. LJ2018B020152) and Chongqing Technology Innovation and Application Demonstration Project (No. cstc2018jscx-msyb0577 and No. cstc2018jscx-msybX0073).

Conflict of interest

None to report.

References

Jiang

Xian

Wang

. China’s response to the rising stroke burden. BMJ. 2019; 364: l879.

Filippidis

Kapsalaki

Patramani

Fountas

. Cerebral venous sinus thrombosis: Review of the demographics, pathophysiology, current diagnosis, and treatment. Neurosurg. Focus. 2009; 27: E3.

Ganesh

Estrada

Yeger

Duffin

Cheng

. A non-invasive magnetic resonance imaging approach for assessment of real-time microcirculation dynamics. Sci Rep. 2017; 7(1): 7468.

Pappu

Lerma

Khraishi

. Brain CT to assess intracranial pressure in patients with traumatic brain injury. J Neuroimaging. 2016; 26: 37-40.

Urback

MacIntosh

Goldstein

. Cerebrovascular reactivity measured by functional magnetic resonance imaging during breath-hold challenge: A systematic review. Neurosci Biobehav R. 2017; 79: 27-47.

Zhang

Medow

Iskandar

Wang

Shokoueinejad

Koueik

, et al. Invasive and noninvasive means of measuring intracranial pressure: A review. Physiol Meas. 2017; 38: R143-R182.

Rasulo

Bertuetti

Robba

Lusenti

Cantoni

Bernini

Girardini

Calza

Piva

Fagoni

, et al. The accuracy of transcranial Doppler in excluding intracranial hypertension following acute brain injury: A multicenter prospective pilot study. Crit Care. 2017; 21(1): 44.

Haq

Mathur

Singh

Kaur

Sibia

Badhan

. Colour doppler evaluation of extracranial carotid artery in patients presenting with acute ischemic stroke and correlation with various risk factors. J Clin Diag Res. 2017; 11(3): TC01-TC05.

Malaeb

Izzetoglu

McGowan

Delivoria-Papadopoulos

. Noninvasive monitoring of brain edema after hypoxia in newborn piglets. Pediatric Research. 2017; 264.

10.

Soldatos

Karakitsos

Chatzimichail

Papathanasiou

Gouliamos

Karabinis

. Optic nerve sonography in the diagnostic evaluation of adult brain injury. Crit Care. 2008; 12(3): R67.

11.

Yang

Dai

Shi

Dong

. A novel multi-frequency electrical impedance tomography spectral imaging algorithm for early stroke detection. Physio Meas. 2016; 37(12): 2317-35.

12.

Aristovich

Packham

Koo

dos Santos

McEvoy

Holder

. Imaging fast electrical activity in the brain with electrical impedance tomography. Neuro Image. 2016; 124: 204-13.

13.

Hart

Meyer

, et al. A noninvasve electromagnetic conductivity sensor for biomedical applications. IEEE Trans Biomed Eng. 1988; 35(12): 1011-22.

14.

Gonzalez

Rubinsky

. Frequency dependence of phase shift in edema: a theoretical study with magnetic induction. In Proceedings of the 2005 IEEE Engineering in Medicine and Biology 27th Annual Conference. 2005.

15.

Yang

Qiao

Wang

. Cardiopulmonary signal detection based on magnetic induction. Journal of Sensors. 2017; 2017: 1-9.

16.

Manoufali

Bialkowski

Mohammed

Mills

Abbosh

. Near-field inductive-coupling link to power a three-dimensional millimeter-size antenna for brain implantable medical devices. IEEE Trans Biomed Engineering. 2018; 65(1): 4-14.

17.

Perron

RRG

Iskander

Seto

Huang

Bibb

. Electromagnetics in Medical Applications: The Cardiopulmonary Stethoscope Journey. The World of Applied Electromagnetics: In Appreciation of Magdy Fahmy Iskander. 2018; 443-79.

18.

Watson

Morris

Williams

Griffiths

Gough

. A primary field compensation scheme for planar array magnetic induction tomography. Physiol Meas. 2004; 25: 271-9.

19.

, et al. Performance of cerebral hemorrhage simulation detection system based on magnetic phase shift spectrum method. Biomedical Engineering & Clinical Medicine. 2011; 15: 505-508.

20.

Sun

Jin

Qin

Wan

Wang

Guo

Ning

, et al. Detection of acute cerebral hemorrhage in rabbits by magnetic induction. Brazilian Journal of Medical and Biological Research. 2014; 47(2): 144-150.

21.

Jin

Sun

Qin

, et al. A new method for detecting cerebral hemorrhage in rabbits by magnetic inductive phase shift. Biosensors and Bioelectronics. 2014; 52: 374-78.

22.

Jin

, et al. A special phase detector for magnetic inductive measurement of cerebral hemorrhage. Plos One. 2014; 9: e97179.

23.

Pan

Yan

Qin

, et al. Detection of cerebral hemorrhage in rabbits by time-difference magnetic inductive phase shift spectroscopy. Plos One. 2015; 10(5): e0128127.

24.

Sun

Jin

Qin

Feng

. Twenty-four-hour real-time continuous monitoring of cerebral edema in rabbits based on a noninvasive and noncontact system of magnetic induction. Sensors. 2017; 17(3): 537.

25.

Sun

Jin

Chen

Qin

Feng

. The experimental study of increased ICP on cerebral hemorrhage rabbits with magnetic induction phase shift method. Iranian Journal of Medical Physics. 2016; 13(2): 125-136.

26.

Yan

Jin

, et al. Magnetic inductive phase shift: A new method to differentiate hemorrhagic stroke from ischemic stroke on rabbit. Biomedical Engineering Online. 2017; 16(2): 63.

A cerebral edema monitoring system based on a new excitation source

Abstract

BACKGROUND:

METHODS:

RESULTS:

CONCLUSION:

Keywords

1. Introduction

2. Methods

2.1 Design of the fixed frequency detection system

2.1.1 Signal source

2.1.3 Sensor

2.2.1 Signal source test

2.3 Statistical analysis

3. Results

3.1 Performance test of the signal source

Table 1 Frequency stability of excitation signal test datum

Table 3 Relationship of MIPS and volume change

5. Conclusion

Footnotes

Acknowledgments

Conflict of interest

References

Table 1
Frequency stability of excitation signal test datum

Table 3
Relationship of MIPS and volume change