2.1.

Structure and Fabrication

The pressure and temperature sensor is essentially an optical fiber with an internal FBG and a reflecting diaphragm. The SMF is a low bend loss fiber with draw tower gratings from FBGs¹⁹ with a cladding diameter of $125\mu \mathrm{m}$ and a mode field diameter of $6\mu \mathrm{m}$, and is encapsulated by a glass capillary (inner diameter of $130\mu \mathrm{m}$; outer diameter of $200\mu \mathrm{m}$). The capillary is fused by a Siecor fusion splicer M90 to the SMF on one side and sealed on the other side with an optical powercore multimode glass fiber (o.d. of $200\mu \mathrm{m}$). The powercore fiber is polished with $0.3\mu \mathrm{m}$ corning-sized diamond polish paper down to $\sim 6-10\mu \mathrm{m}$. Afterward, the diaphragm thickness is reduced by hydrofluoric (HF)-acid to $h\sim 2\mu \mathrm{m}$.²⁰ This construction forms a pressure sensitive three-mirror EFPI with a very thin diaphragm, as shown in Fig. 1(a). After the etching process, the pressure sensors exhibit a stability of better than $1\mathrm{cm}{\mathrm{H}}_2\mathrm{O}/\text{hour}$ with a resolution of more than $0.1\mathrm{cm}{\mathrm{H}}_2\mathrm{O}$, tested in a 60 cm burette filled with water. The FBG is inscribed in the core of the SMF in close proximity to the point of measurement of the EFPI. The light travels from a broadband light source (BLS) to the end of the fiber-tip [i.e., the end of the sensor, as shown in Fig. 1(a)]. The reflected signal returns to an optical spectrum analyzer (OSA). The light from the FBG is partially reflected before reaching the tip of the fiber. This results in a high intensity of reflection at a particular wavelength, depending only on temperature and strain.²¹ The narrow band reflection is shown in Fig. 1(b) by a sharp peak in the 1550 nm wavelength range.

Fig. 1

(a) The optical fiber pressure and temperature sensor (OFPTS) based on a single-mode fiber (SMF) with internal fiber Bragg Grating (FBG). (b) OFPTS spectrum: illustrating the superposition of the FBG peak and the broadband Fabry–Perot interferometer (EFPI) spectrum.

Since the sensor is made of glass only, sensitivity parameters are determined by the Youngs modulus ($E$) and the Poissons ratio ($\mu$), as described in Eq. (1). The structure of the sensor allows the end mounted diaphragm to bend (i.e., changing the cavity length $\mathrm{\Delta }G$) by

2.2.

Optical Fiber Pressure and Temperature Sensor Placed in a Medical Catheter

The pressure sensor is placed at the end of a 1 m optical fiber. The bend insensitive fiber combined with additional algorithms supports a bend radius of less than 5 mm. The end of the fiber is spliced to a buffered and jacketed 5 m SMF-28 fiber and terminated using an FC- angle polished connector (FC-APC). The sensors are guided through a Y-junction as shown in Fig. 2(a) up to the tip of the catheter. Two OFPTSs were placed into a 5 Fr nutrisafe 2 catheter (REF 361.052) from Vygon with a length of 50 cm. The fiber is externally glued to the outside of the Y-junction to guarantee a stable placement (with no interference of glue and saline solution) of the tip of the sensor between hole one and two and behind hole two and three in a 1 cm distance. The glue also seals the end of the catheter completely, which is mandatory for an accurate measurement.

Fig. 2

(a) The schematic diagram of the OFPTS in a medical catheter. Both sensors are placed at a 1 cm distance to each other and sealed by a Y-junction. (b) The photograph of both sensors illuminated by a red light to identify the position in the catheter.

The Y-junction also allows the infusion of 0.9% saline solution to remove air from the catheter. This prevents a rapid change of the refractive index from saline solution ($n\sim 1.34$) to air ($n\sim 1$). The whole catheter-sensor structure can be easily replaced by connecting two new sensors to the FC-APC of the system. Figure 2(b) photographically shows the catheter with sensors used during the uryodynamic measurement.

2.3.

Optical Urodynamic System

The optical urodynamic system (OUS) is shown in Fig. 3(a). The BLS (EXS210069-01) from Exalos has a Gaussian shape output with a power of 14.97 mW and a bandwidth of 46.2 nm at 1554.2 nm. The amplified spontaneous emission ripple is quoted as 0.074 dB over a wavelength range of 0.5 nm. The source is connected to a 3-dB splitter guiding the light to an optical switch from Sercalo (SW1x4-9N) with a 1 ms switching time and a 1.0 dB insertion loss, based on micromechanical mirrors. The light is reflected (at the FBG and EFPI) and guided back to the OSA (IMON-512 from Ibsen Photonics with a bandwidth from 1510 to 1596 nm). The OSA has a resolution of 512 pixels corresponding to a peak shift of the EFPI spectrum of 0.166 nm. The relatively low spectral resolution is compensated by algorithms developed for EFPI pressure measurements, which also compensate for any signal artifacts and rapid changes of saline concentration. The optical switch was operated with a frequency of 20 Hz in an interval of $5\text{frames}/\text{cycle}$ and $2\text{cycles}/\text{sensor}$ resulting in a 10 Hz sample rate for each sensor, as recommended for urodynamic analysis.² An Arduino one board triggered the optical switch and activated the OSA with a time delay of 2 ms to allow the optical switch to stabilize. The signal was acquired for $20\mu \mathrm{s}$, however, this could be varied depending on the reflectivity of the sensor (i.e., weaker gratings and higher loss in splicing need a higher acquisition time). The complete system is housed in a hand case for protection and ease of transport.

Fig. 3

(a) The optical urodynamic system (OUS), based on a light source, 3-dB coupler, Arduino one board, an optical switch (4 ports) and an optical spectrum analyzer (OSA). Everything fits in a hand case for easy transport and protection. (b) The graphical user interface (GUI) of the urinary measurement. The first graph shows the pressure trace of the first OFS in the tip ($P_{{\mathrm{OFS}}_1}$). The second graph shows the pressure of the second OFS ($P_{{\mathrm{OFS}}_2}$) and the third graph shows the differential pressure ($P_{\mathrm{diff}}$) between the first and second sensors.

The software is based on a three-layer concept. First, the synchronization of the switch and optical spectrum analyzer was timed by an Arduino one board programmed in C. The OSA acquired the light from the sensor and buffered the frames internally in an internal buffer. Second, a server programed in C++ communicates to the interrogator and provides the frames (i.e., measurement of the intensity of the reflected spectrum) as packages to a client over a network protocol. Finally, the client was an in-house developed LabView program designed for urodynamic analyses. The graphical user interface is shown in Fig. 3(b) and shows three main graphs, indicating the pressure trace of both OFPTSs and a third graph for differential pressure. The software also compensates for small bending and changes in the refractive index of the saline medium. An internal additional algorithm was developed to overcome the relatively low spectral resolution of the interrogator and to achieve a high resolution of better than $0.1\mathrm{cm}{\mathrm{H}}_2\mathrm{O}$.¹³

3.1.

Setup of the Equipment

In Fig. 4, the OUS is shown on the left-hand side, whereas the medical equipment is schematically shown on the right-hand side. In each patient, two commercially available medical sensors were placed in the bladder and in the rectum, respectively. A separate catheter was placed in the bladder to facilitate the filling phase with saline solution and a dynamic flowmeter was used to analyze the flow rate of urine during voiding.

Fig. 4

The figure shows the schematic of the optical and standard medical equipment. From the optical urodynamic system, one catheter with two sensors was inserted. The medical equipment (DYN) consists of two separate catheters for each sensor (middle catheter) and an additional catheter (on the top) to fill the bladder with saline solution. During the urination process, a flowmeter was used to analyze the urinary flow rate.

3.2.

Medical Reference Sensor: PICO-2000 and SmartDyn

To guarantee an appropriate measurement comparable with current gold standard urodynamic systems [Figs. 5(a) and 5(b)], two commercial systems were used and the results were compared with the OUS under identical conditions. The PICO-2000 is a urodynamic system produced by MenFis bioMedica.²⁴ The second system provided by Albyn Medical Ltd. is called SmartDyn. Both systems are based on electronic sensors, measuring a differential pressure in the bladder (i.e., the intravesical pressure $P_{\mathrm{ves}}$) and the abdomen (i.e., the intra-abdominal pressure $P_{\mathrm{abd}}$). The SmartDyn system is available with a main control which connects the sensor, flowmeter, and infusion system. The flowmeter is based on a gravimetric measurement whereby the patient urinates into a beaker placed under a chair. This is used to analyze the speed and volume during urination (dynamic flow analysis). Additionally, a controlling system infuses saline solution through the urethra into the bladder of the patient. All measurements are collected and transferred to a PC. Figure 5(a) shows the desk with the laptops and the OUS in the background. The first laptop was used to analyze the OFPTS, the second laptop was wirelessly connected to the SmartDyn system.

Fig. 5

(a) The measurement of the urination process is simultaneously analyzed on each laptop. (b) The examination room is separated by a curtain in two areas. The main area contains the examination table where the patient is placed during the insertion of the sensor.

3.3.

Examination and Preparation

To analyze lower urinary tract symptoms (LUTSs) (e.g., incontinence, urgency to urinate),²³ the patient needs to undergo a cystometrogram (CMG) which involves filling the bladder with saline solution ($\sim 0.9\%$ salt concentration). In a healthy bladder, the result is a minimal increase in pressure, however, if the bladder pressure increases beyond a determined point, this indicates that the bladder is unhealthy and lacks elasticity (known as bladder compliance).²⁵ The pressure–volume relation depends on various factors, e.g., gender, shape, and volume of the bladder.²⁶ Furthermore, dynamic pressure responses caused by coughing or talking may simultaneously affect the bladder and abdomen. When the patient starts to urinate the detrusor muscle applies pressure mainly on the bladder. This can be expressed in

The complete process of filling the bladder through voiding is generally measured in vivo using pressure probes in both the bladder and rectum. A pressure-flow study measures the pressure of the bladder with additional analysis of the flow rate as measured by a flowmeter.²³ The pressure measurement is recorded by two sensors, one placed in the bladder and the other in the rectum. In the work presented here, an alternate method is investigated, whereby two sensors have been inserted in the same urethra catheter at a distance of 1 cm apart. The first sensor is placed in the bladder (${\mathrm{OFS}}_1$) and the second close to the urethra (${\mathrm{OFS}}_2$). The distance of 1 cm was chosen based on the functional profile length over the maximum urethral pressure.²⁷ The highest pressure change ($\mathrm{\Delta }P$) can be observed over a 1 cm distance, which can be used as an indicator for correct placement. During the insertion process, when both sensors show a similar pressure measurement (i.e., $P_{\mathrm{diff}}$ is comparatively small), both sensors are correctly placed. The differential pressure ($P_{\mathrm{diff}}$) is calculated by

Before the examination starts, the patient is placed on the examination table [Fig. 5(b)] and both commercial sensors in the catheters are guided through (1) the urethra into the bladder and (2) into the rectum of the patient. The catheter with both OFPTSs is inserted afterward, as shown schematically in Fig. 6. The entire length of the catheter was inserted deep inside the bladder to ensure that both OFPTSs were placed well inside the lumen (cavity). In this way, both sensors were able to record the same pressure, regardless of the variable patient anatomy. After this, the patient was made to cough and if both optical sensors were approximately sensitive to the cough, it was considered that the sensors were fully inside the urethra.

Fig. 6

Anatomy of female patient.

The urodynamic examination is divided into two phases. In the first phase, the patient is asked to cough several times to test all sensors and the bladder is infused with saline solution (CMG). At the end of this phase, the bladder reaches a certain maximum of volume capacity at which point the patient starts to feel the urge to empty the bladder which indicates that the bladder is fully contracted. The patient then moves to the urination chair for the voiding phase (pressure flow study). Since the sensors are moving and bending during this time, the equipment may have to be moved or even disconnected rendering the pressure readings irrelevant. In the second phase, the patient empties the bladder into a receptacle with a flowmeter situated below recording the volume of urine. In a healthy patient, the muscles contract and squeeze the urine out of the bladder during urination. In a patient who exhibits a urological disorder this function may be impaired or may not work at all.

4.1.

Feasibility Test with Bend-Sensitive Fiber-Based Sensor

To investigate the feasibility of an optical fiber sensor in a urodynamic measurement, a single OFPTS was tested in vivo in the bladder of a female patient. The commercial medical sensor, used as reference sensor in this measurement, was the PICO-2000. In the first part, the bladder is infused with saline solution, which results in an increase of the bladder pressure ($P_{\mathrm{ves}}$). Both sensors show a very good correlation during the infusion, with a correlation factor of $R^2=0.93$, as shown in Fig. 7. As soon the bladder contraction starts, the patient is moving from the examination bed to the urodynamic chair, causing a high bend in the bend sensitive fiber. The pressure trace is still observable but the correlation factor of both sensors is reduced to $R^2=0.01$. In the last phase, the patient is in the voiding phase. The bladder contractions are measured by both sensors with a high correlation of $R^2=0.81$.

Fig. 7

The comparison of the PICO-2000 and the OFPTS shows a good correlation. During the movement of the patient, the bend sensitive fiber lost the signal, causing an offset in the measurement.

Since the fiber is highly sensitive to bending which can cause a variation in the reflected signal, the online interpretation of the data during the 16 min in vivo analyses was inaccurate. However, in postanalysis, the high degree of correlation of both sensors became evident.

4.2.

High Sensitivity with Bend-Insensitive Sensor

The sensors were initially built by a corning optical fiber (SMF-28). Further investigation demonstrated that the bending during the examination changed their reflected spectrum with a bend radius below $\sim 2\mathrm{cm}$, causing artifacts which affect the online measurement. As a result, the SMF of the OFPTS was replaced with bend insensitive fibers from FBGS. Furthermore, the PICO-2000 system was changed to a more advanced medical urination system (i.e., SmartDyn). Figure 8 shows the recoded pressure data from an online measurement, comparing the OFPTS with the SmartDyn sensor.

Fig. 8

A 5-min measurement of the patient with infusion of saline solution in (A) results in a variation of pressure. (a) A 1-min measurement during cough and infusion of the SmartDyn system. (b) The same time period as (a) with the OFPTS.

The patient’s cough in instances (B) have both similar values, indicating a good correlation. The period covered by part (A) shows when the bladder was being filled with 200 ml of saline solution. The volume of the infusion is controlled and recorded by the SmartDyn system. The SmartDyn sensor shows a pressure increase of $\sim 10\mathrm{cm}{\mathrm{H}}_2\mathrm{O}$, whereas the OFPTS increase $\sim 20\mathrm{cm}{\mathrm{H}}_2\mathrm{O}$. Even with an initialized time resolution corresponding to a sampling frequency of 10 Hz, the details of the OFPTS are clearer than those of the medical sensor. The peak in (B) measured by the OFPTS in Fig. 8(b) shows two peaks per cough. The SmartDyn sensor in Fig. 8(a) shows an envelope of both peaks. This explains the smooth signal of the SmartDyn sensor during the measurement. The measurement of the OFPTS is shown without additional filtering of the pressure measurement, resulting in a more detailed observation.

4.3.

Dual Optical Fiber Pressure and Temperature Sensors System and the Medical Sensor

The previous analysis demonstrated a good correlation between OFPTS and the commercially available sensors. The high sensitivity of the sensor was capable of detecting even small variations during the saline infusion. Investigation into the reliability of the OUS has to be proved by a simultaneous measurement of all four sensors during a full investigation. An additional examination with both optical sensors (${\mathrm{OFS}}_1$ and ${\mathrm{OFS}}_2$) in a 5 Fr catheter was used in a full analysis alongside the bladder (ves) and abdominal (abd) sensors. The results from filling the bladder up to the voiding phase are shown in the 14 min measurement in Fig. 9.

Fig. 9

Pressure data captured using both OFPTS (i.e., the first in front of the catheter and the second 1 cm behind the first) and the SmartDyn sensors in the bladder and abdomen.

In urodynamics, the morphology of the contraction caused by the bladder muscles is important and requires a high precision measurement. Significantly, part (B) shows contraction during the infusion of saline solution with the same pattern in all four sensors. However, only the two OFPTSs show a significant increase in pressure during the infusion time, such as would be the case for a normal healthy bladder.²⁵ Parts (C) and (D) show a series of coughs and (F) shows the patient’s movement, which are not relevant for this urodynamic analysis. (E) shows some contraction of the muscles of the bladder, indicating the urgent need to urinate. The pressure measurement demonstrates that all four sensors are measuring similar pressures, indicating a good correlation.

5.1.

Complete Differential Pressure Analysis

Urodynamic studies are mainly interested in the contraction of the bladder which is generated by the bladder muscles. Other bodily movements such as coughing and talking can affect the pressures generated in the bladder, abdomen, and rectum. To remove this artifact, the differential pressure analysis is calculated [see Eqs. (2) and (3)]. Figure 10(a) shows the pressure measurement of both optical sensors ($P_{{\mathrm{OFS}}_1}$ and $P_{{\mathrm{OFS}}_2}$). The bladder pressure ($P_{\mathrm{ves}}$) and abdominal pressure ($P_{\mathrm{abd}}$) are recorded in Fig. 10(b). Figure 10(c) shows the differential pressure measurement of both techniques. The infusion of saline solution and the voiding volume measured by the flowmeter are shown in Fig. 10(d). The cystometric capacity (CC) indicates a full bladder which results in a higher activity in the pressure change.²

Fig. 10

(a) The pressure measurement by both OFPTSs. (b) The pressure by the commercial medical reference sensor. (c) Compares the differential pressure of the OUS and the SmartDyn system. (d) The volume of the infusion of saline solution and the volume measured by the flowmeter during voiding the bladder. The cystometric capacity (CC) is reached at 250 ml after the cough, resulting in a higher activity in the pressure change.

During the time of 1 to 11 min, the bladder is filled with saline solution which results in an increase of the volume, as shown by $V_{\inf }$ in Fig. 10(d). Only the first OFPTS ($P_{\mathrm{OFS}1}$) is measuring a rise in pressure. In this case, the commercial abdominal sensor measured a pressure increase, resulting in a negative differential pressure, when it was anticipated that the differential pressure would be positive due to the increase in volume. On the other hand, the differential measurement of the OUS shows an increase in pressure. This can be explained by the high sensitivity of the optical sensors. In our investigation, ${\mathrm{OFS}}_1$ was fully inside in the bladder, whereas the second sensor (${\mathrm{OFS}}_2$) was close to the patient’s urethra. This resulted in the same pressure reading when muscle contraction occurred. The first sensor also measured the increase in pressure caused by the increase of volume in the bladder during infusion. Indeed, all four sensors show the same trend, but only the first optical sensor (${\mathrm{OFS}}_1$) shows an increase in pressure with the infusion of saline solution and stays constant after the infusion has stopped.

In the second part (B) of the measurement, the patient started to urinate. The increase in $V_{\mathrm{uri}}$ in Fig. 10(d) indicates the volume decrease in the bladder as recorded by the flowmeter. During the urination phase, the correlation of both sensors is evident. Furthermore, the differential pressure analysis demonstrates that the small variations are caused by the obstruction since they occur on both optical sensors simultaneously and are completely removed.

5.2.

Pressure Performance During Infusion and Voiding

A detailed analysis is shown in Fig. 11 which shows the bladder contraction of the patient (${\mathrm{PAT}}_1$) directly after the CC from the measurement in Fig. 10. Both sensors (${\mathrm{OFS}}_1$ and ${\mathrm{OFS}}_2$) show a good similarity, which demonstrates the high sensitivity of the sensors. To compare the structure of the obstruction, a second patient (${\mathrm{PAT}}_2$) with a clear contraction trend is also shown on the normalized time graph. Unlike the first patient where small variations in the graph were visible, in the case of ${\mathrm{PAT}}_2$ a normal graph was demonstrated. This could be an indicator that ${\mathrm{PAT}}_1$ has a urological condition.²

Fig. 11

Bladder contraction of patient 1 (${\mathrm{OFS}}_1$ and ${\mathrm{OFS}}_2$), with changed pressure offset (to compare both charts). The bladder contraction of a second patient (${\mathrm{PAT}}_2$) is shown on the normalized time graph to compare the obstruction pattern.

The infusion of saline solution (i.e., increasing of bladder volume) is controlled by the SmartDyn system. This results in an increase in the bladder pressure ($P_{{\mathrm{OFS}}_1}$) only. However, the voiding of the bladder has to be controlled by the detrusor muscle, which effects both sensors. The differential pressure ($P_{\mathrm{diff}}$) as a function of the bladder pressure ($P_{{\mathrm{OFS}}_1}$) is shown in Fig. 12(a). The figure is divided into an infusion and a urination (voiding) phase. It is clearly evident that during infusion, only the pressure of the optical sensor in the bladder ($P_{{\mathrm{OFS}}_1}$) has increased. During the voiding (emptying) phase, the slope is smaller ($m=0.43$) since the differential pressure ($P_{\mathrm{diff}}$) during the change of bladder pressure ($P_{{\mathrm{OFS}}_1}$) has become smaller. This can be explained by the fact that the detrusor muscle is affecting the bladder and the urethra (i.e., $P_{{\mathrm{OFS}}_1}$ and $P_{{\mathrm{OFS}}_2}$), resulting in a reduced differential pressure.

Fig. 12

(a) Pressure relation of differential pressure $P_{\mathrm{diff}}$ in relation to the bladder pressure $P_{{\mathrm{OFS}}_1}$ during infusion and urination phase. The slope (m) of the linear fit (LF) is plotted for both phases. (b) The volume of the bladder ($V_{\text{Bladder}}$) as a function of the differential pressure ($P_{\mathrm{diff}}$). The infusion is shown on the left-hand side and the urination on the right-hand side.

The function of the differential pressure in relation to the volume after the CC was reached is shown in Fig. 12(b). This demonstrates that the differential pressure is linearly increasing during the infusion with $0.1\mathrm{cm}{\mathrm{H}}_2\mathrm{O}/\mathrm{ml}$. The same behavior was also demonstrated by Malbrain and Deeren,²⁸ whereas the voiding phase shows a bigger change in differential pressure ($P_{\mathrm{diff}}$) with the change of bladder volume ($V_{\text{Bladder}}$). In the research by Malbrain and Deeren, the infusion is on the right-hand side and the voiding phase is on the left-hand side (i.e., the opposite of this measurement). This could be an indication of a medical condition.²⁸