An Experimental Study on Bipolar Tissue Hemostasis and Its Dynamic Impedance

Bipolar tissue hemostasis is a medical procedure where high frequency alternating current is applied to biological tissue for wound closing and blood vessel sealing through heating [1–6]. The mechanism of hemostasis using bipolar heating is the denaturation of protein in the blood vessel wall to cause fusion of the vessel. From a manufacturing point of view, the process is similar to resistance spot welding that has been extensively studied [7–9]. Both processes utilize resistive heat to join the material clamped in between two electrodes. Bipolar tissue hemostasis has been used in many medical specialties including gynecologic, urologic, cardio-thoracic, and colorectal surgery [10]. Despite the advantages of fast operation, shorter recovery time, and suitability for minimal invasive operation, problems such as tissue sticking, peripheral thermal damage, and postprocedure seal failure are still major concerns [11–21].

A number of studies have been conducted to address the problems in bipolar hemostasis. Reyes et al. [22] performed a blood vessel sealing experiment to identify the best process conditions for achieving the highest postweld burst pressure. Anderson et al. [23] and Chen et al. [11] studied the effect of clamping force during the sealing process. Li et al. [24,25] examined the effect of compression uniformity on the hemostasis process. To reduce tissue sticking and charring, which render the procedure time-consuming and unsafe [26], various surface coatings have been applied to bipolar forceps. Ceviker et al. [27] tested commercially available bipolar forceps with striped Teflon coating. It helped reduce sticking; however, Teflon coating during heating could release toxic particles [28]. Mikami et al. [29] tested the sticking effect of three different types of electrodes, including titanium, stainless steel, and gold plated stainless steel, and found that gold plated bipolar forceps had the least sticking for the same surface roughness.

An important variable in the tissue hemostasis process is electrical impedance, which varies during the process due to the water content and cell structural change. Equivalent to the dynamic resistance in resistance spot welding [30–32], this dynamic impedance has been used as a signal to control the bipolar hemostasis process. Vällfors and Bergdahl [33,34] suggested that the dynamic impedance during a bipolar vessel sealing process could be divided into several stages and tissue coagulation formed when the impedance reached the minimum. This understanding was used to design a control algorithm to stop the heating power once the impedance reached a threshold [35]. The initial value of dynamic impedance was also used to determine the amount of energy to deliver in a vessel sealing device [36]. However, there still lacks a reliable method to ensure the seal quality and avoid tissue sticking and charring.

In this research, we conduct a systematic study to examine process parameter effects in the bipolar tissue hemostasis process. Heating voltage and current are measured to determine the dynamic tissue impedance during the process. Tissue sticking levels are defined and correlated to the denatured tissue zone size. The effects of electrode compression level, heating power, and time are analyzed. Heating energy and bio-impedance are evaluated for quality prediction. This study lays ground for developing an online monitoring and control method for quality assurance for the bipolar tissue hemostasis process.

Experiments were conducted to simulate an actual hemostasis surgical procedure. The samples used were excised porcine muscle tissue, because of its similarity to human tissue. Porcine tissue was acquired from a local supermarket. It was first frozen, sliced into 20 × 20 × 5 mm samples, and thawed to room temperature before experiments. The cut samples were wrapped in aluminum foil to preserve water content while thawing. Before each test, excess water on the surface of tissue samples was removed with absorbent wipes.

Figure 1(a) shows a schematic of the experimental setup. A bipolar electrosurgical generator (ConMed Corporation, Sabre 2400, Utica, NY) was used to provide power for tissue coagulation. The forceps used was a commercial laparoscopic hand-piece (Therma-Sect 010713-01, Karl Storz GmbH & Co. KG, Tuttlingen, Germany). The heating current was measured using a toroid current sensor made in-house. The voltage was measured with lead wires. The heating voltage and current were used to determine the electrical impedance of tissue during the coagulation process. The power generator used in this study had a pulse frequency of 20 kHz. Within each pulse, the electrical power delivered was a 1 MHz sinusoid. The pulse width was modulated to control the power output. Figure 1(b) shows the raw voltage signal captured for the first 2 ms under the 35 W power setting. The current signal has the same pattern but different magnitude. To reduce the sampling rate requirement, both voltage and current signals were rectified with a diode and filtered using a low-pass filter with a cut-off frequency of 20 kHz. The signals were then sampled at 250 kHz with a Tektronix digital oscilloscope (MDO 3014, Beaverton, OR). The data were then transferred to a computer for analysis.

Experiments were conducted under different power, time, and compression level conditions. The compression level is defined as the thickness decrease under compression as a ratio of the initial tissue thickness. The laparoscopic hand piece had a locking mechanism, which was used to fix the compression level. Images were taken during and after the heating process to examine the size of denatured tissue zone from both the top surface and cross sections of the sample. A sliding-level experiment [37] was conducted, with compression level and heating power chosen at fixed levels and heating time in a sliding range determined by the other two variables. Table 1 shows the experimental conditions used in this study. For each power and compression condition, at least four different time settings were used.

where $Â$ and $θ̂$ are estimated as the amplitude and phase angle of the signal, $f$ is the frequency, k is the digital sample number, and N is the number of total digital samples of the signal. Since the voltage and current signal varied during the heating process, the estimation was repeated every 1000 data points, which corresponds to a window of 4 ms. Within this short window, the voltage and current are assumed to be constant.

where T is the duration of the heating process.

In this study, the quality of tissue hemostasis process is defined as the size of denatured tissue zone. The first sign of tissue denaturation is the change in color. The denatured tissue zone was visually identified by color difference and measured using the image processing software imagej. Figure 2 shows an image of the heated tissue sample. The inner dash line shows the size of the electrodes and the outer dash line shows the size of the denatured tissue zone. In this study, we use a relative denatured zone size (S_r) as the quality measure. S_r is defined as the ratio of the denatured tissue zone size, circled in solid lines in Fig. 2, to that of the jaw electrodes, which was 55 mm² in this study. The size of the electrode is marked using a dashed line in the figure. The software imagej is used to find the areas based on the marked image. The tissue sticking level is also analyzed as a response variable. The sticking level is subjectively defined based on the easiness to remove tissue residue from the teeth of the electrodes after each weld. Table 2 shows the definitions of sticking levels. The cleaning processes were all performed by the same operator to avoid bias error.

Tests under 25 W power setting and 50% compression level were conducted from 1 to 4 s to examine the growth of denatured tissue zone in the bipolar hemostasis process. Both top surfaces and cross sections of samples were examined. The results are summarized in Table 3. Denatured tissue becomes white in color. This visual inspection has been used by surgeons to determine when to terminate the heating process. From the top surface images, it can be seen that tissue started to denature from the proximal end of electrode jaws. From the cross-sectional images, especially Section A at 4 s, it can be seen that tissue started to denature in the center of each cross section. Once the top surface turns white, it is certain that the tissue underneath is fully denatured. Therefore, the denatured tissue zone as seen from the top surface is used as an indicator of joint quality in this study.

The effect of heating process parameters on the size of denatured tissue zone is shown in Fig. 3. The relative denatured zone size S_r at different conditions is plotted against heating time. Each of Figs. 3(a)–3(c) shows results with the same heating power, but different compression levels. It can be seen that the denatured tissue zone size linearly increases with the heating time. However, the line shifts upward when a higher compression level is used. Under the same power and heating time, a higher compression level would lead to a larger denatured tissue zone. This may be due to the fact that with a higher compression level, more tissue is pushed aside and a thinner target tissue is left between the electrodes. While the denatured tissue zone on the top surface is larger, the total denatured tissue volume could remain the same. Figures 3(d)–3(f) show results with the same compression level, but different heating power. Under the same compression level and heating time, a higher power will produce a larger denatured tissue zone, as expected.

Since power and heating time are the two major process parameters in the bipolar hemostasis process, a process map that is similar to a weld lobe for resistance spot welding can be constructed with the above experimental data. Figure 4 shows such a lobe that defines under heated, acceptable, and overheated regions in a power-time plane. The lower and upper bounds of the acceptable region can be defined according to user requirements. In Fig. 4, they are chosen to be S_r = 1 and S_r = 2 for illustration. A weld lobe is generated for each of the compression levels using exponential curve fitting. From locations of these three lobes, it is easy to see that a shorter heating time can be used to achieve an acceptable denatured tissue zone when the compression level is higher. However, a lower compression level in general has a wider acceptable lobe region, which means that the process is more tolerant under this condition. With the same power setting, a lower compression level allows a longer time span between the lower and upper bounds to achieve an acceptable result.

Heating energy is examined for predicting the denatured tissue zone size under various process conditions. Figure 5 shows plots of S_r as a function of heating energy under different conditions. The heating energy was determined based on Eq. (7) with measured welding voltage and current. It can be seen in Fig. 5 that compression level and power setting have different effects on the quality of the bipolar hemostasis process. Under the same compression level, the denatured tissue zone size could be well predicted with the heating energy, regardless of the heat power used. This can be seen in Figs. 5(a)–5(c). There is a linear relationship between the denatured tissue zone size and the heating energy. Under different compression levels, the same heating energy will not yield the same denatured tissue zone size, as shown in Figs. 5(d)–5(f). With the same power setting, a higher compression level generally yields a larger denatured tissue zone with the same heating energy used.

As discussed above, a higher compression level may lead to more tissue being pushed aside and thinner target tissue left for heating. This could result in a larger denatured tissue zone observed from the top surface, but the same volume under the same heating energy. We therefore examine the denatured tissue volume as a function of the heating energy under different compression levels. The denatured tissue volume was calculated by multiplying the size of the denatured tissue on the surface of the sample by its corresponding compressed tissue thickness. This result is shown in Fig. 6. It can be seen from the figure, although the volume of denatured tissue zone appears to be linearly related to the heating energy, this relationship is still strongly affected by compression level. Under the same heating energy, a lower compression level leads to a larger denatured tissue volume.

Bioimpedance during heating was determined to understand the dynamic behavior of the tissue hemostasis process. Figure 7 shows a typical dynamic impedance curve for tissue samples heated at 25 W and 75% compression level. Samples with different heating durations are also shown in the figure. It can be seen that, in general, the dynamic impedance curve can be divided into two regions. Initially, the impedance will decrease due to the increase of contacting area between the electrodes and tissue caused by the thermal expansion of porcine tissue under heating. As the heating process continues, the desiccation process starts, causing the impedance to increase due to the loss of water. The same observation has been discussed by Bergdahl and Vällfors [33]. However, the dynamic impedance was not related to the denatured tissue zone size. From the correlation between the dynamic impedance and the denatured tissue zone images, it can be seen that the denatured tissue zone almost achieved the size of electrode when the dynamic impedance reached the minimum. No sticking was observed at this moment.

Figure 8 shows dynamic impedance curves under different compression and power settings. It can be seen from Figs. 8(a)–8(c) that a higher compression level causes a lower impedance. This is due to the larger contact area and shorter electrical path achieved under high compression. With the same power setting, a higher compression level tends to result in a higher desiccation rate after the minimum impedance is achieved. Figures 8(d)–8(f) show that with the same compression level, a higher power setting will cause a faster heating and desiccation process of the tissue.

Dynamic impedance has been used to control the bipolar hemostasis process. Vällfors and Bergdahl [34] suggested that the heating power should be terminated “soon after” the impedance reached its minimal value. However, the exact criterion on how soon the power should be terminated was not specified. In our experiment, we examine the time that the impedance reaches the minimum and its corresponding denatured tissue zone size. The result is shown in Table 4. It can be seen that depending on the power and compression level settings, the relative denaturation zone size (S_r) varied from 0.93 to 1.2 when the minimal impedance is achieved. This indicates that the minimal impedance criterion alone may not be able to provide consistent prediction of weld quality.

Figure 9 examines the relationship between sticking level and denatured tissue zone size under different power and compression-level conditions. In general, a larger denatured tissue zone is accompanied by a higher sticking level. At the same power setting, a higher compression level caused more sticking for the same denatured tissue zone size. With the same compression level, a higher power setting caused more sticking for the same denatured tissue zone size. These characteristics are more pronounced at high power and high compression level settings, as shown in Figs. 9(c) and 9(d). This result suggests that, for the same denatured tissue zone size, a low power and low compression level setting should be used to avoid sticking.

Experiments on bipolar tissue hemostasis were conducted under various compression, heating power, and time conditions. The dynamic electrical impedance was estimated based on in-process measurements of voltage and current. Effects of process parameters on the size of denatured tissue zone, dynamic impedance, and tissue sticking were analyzed. It is found that tissue denaturation starts from the center of the heated region. The size of denatured tissue zone could be predicted with the heating energy; however, this relationship is strongly dependent on the compression level. The dynamic impedance can be used to indicate the stages of tissue hemostasis process. It is also strongly affected by the compression level, as well as the heating power. Existing criteria for quality prediction and control using heating energy and minimal impedance are not reliable. Impact of compression level and power setting should be considered. To avoid sticking, a low power and low compression level setting should be used for the same denatured tissue zone size. This study lays groundwork for future development of a robust monitoring method for hemostasis quality prediction.

• Directorate for Engineering (Grant Nos. CMMI-1434584 and CMMI-1642565).