A Soft Resistive Sensor with a Semicircular Cross-Sectional Channel for Soft Cardiac Catheter Ablation

Figures

• 

  Figure 1. A 2D rendered image of a soft linear actuator intended for cardiac catheter ablation described in [18]. The soft device consists of a linear hydraulic actuator that can be pressurized using inert saline solution pumped in through the catheter line. The fibers wrapped around the actuator allow for only one degree of freedom when pressured, so the device can extend and retract perpendicular to the heart wall. The soft sensor that will be expanded on in this paper will be housed in line with the ablation tip in order to accurately output the compressive force experienced by the ablation tip during contact with cardiac tissue. The electrodes and wiring normally associated with radio-frequency ablation is not shown in this theoretical portrayal.

• 

  Figure 2. (A) A 3D rendered image of the soft microsensor design based on the novel fabrication method created by Parekh et al. [19] and spiral design featured in [15]. The sensor exhibits a semicircular channel cross section due to the resulting geometry of 3D printing liquid metal onto cured elastomer. (B) The semicircular channel is embedded 0.5 mm beneath the surface of the elastomeric shell of the sensor. The channel forms a spiral structure to maximize the length and resolution of the sensor.

• 

  Figure 3. (A) A 0.05 mm triangular mesh size was used to characterize the soft sensor using finite element analysis via Matlab’s FEAtool application. (B) A 2D graphical depiction of the mesh point locations that were used to model the deformation of the liquid metal channel. The first recorded point begins at the leftmost point of the spiral structure and follows the arrows around the structure until it reaches the final mesh point at the bottommost point of the channel. (C) Areas outlined by the color-coded squares represent the area under compression during pressure directionality testing.

• 

  Figure 4. (A) The initial and deformation after 0.2 and 0.4 N channel coordinates are plotted above. The channel experienced an average Y displacement of 0.271 mm and 0.542 for the 0.2 and 0.4 N cases, respectively. (B) The zx-plane view of (A). The change in length of the channel is subtle near the center of the sensor but becomes more pronounced as the channel expands distally. The total length change for the 0.2 and 0.4 N cases is 4.48 and 8.97 mm, respectively.

• 

  Figure 5. (A) Resistance vs. force relationships found through FEA modeling and analytical approaches. Matlab’s curve fitting application was used to create the linear approximation, with the equation of the line displayed on the bottom right-hand side. The range of highest ablation success rate (0.1–0.2 N) displays high agreement between the two models. (B) The resistance vs. channel length trends for both the analytical and FEA models are plotted above. The FEA model values are based on applied force values (starting left, going right) of 0, 0.2, and 0.4 N.

• 

  Figure 6. FEA simulation showing increase in the deformation area under compression applied to a central, square area. The figure on the left corresponds to compression with a central square of 0.8 by 0.8 mm in size, whereas the right figure shows compression with a 1.2 by 1.2 mm central square. Each simulation’s data was exported and used to calculate the resistance across the sensor. A parametric sweep of the surface area was simulated starting at a 0.8 by 0.8 mm square area and ending at a 2 by 2 mm square area.

• 

  Figure 7. The resistance as a function of expanding deformation area was determined. An expanding square compression (see Figure 6) at a constant 0.4 N force was simulated using a parametric sweep of the square’s side length. A quadratic function was fit to the to the data points, with the equation of the line displayed.

• 

  Figure 8. (A) Small central compressions displayed an exponential increase in resistance of central line segments. As in Figure 3B, the sensor channel is made up of 324 line segments, with the center being composed of segments 150–170. The figure above displays the resistance generated per segment. As expected, the resistance per segment drastically increased in the center of the sensor when a concentrated pressure was applied. As the central compression area increased, the resistance per segment decreased towards the center of the sensor decreased. (B) The above figure represents Figure 3B overlaid with the compression area used in each simulation trial. Each color of the square contacts corresponds to the curves plotted in (A).

• 

  Figure 9. (A) A free body diagram displaying the conditions in which the lateral pressure sweep was executed. A constant force of 0.4 N was dispersed between the rightmost boundary of the sensor, and one of the X-positions shown in (B). This sweep represents the conditions the sensor would be under if there was an imperfect uniliateral contact of the ablation tip with the heart wall. Having the relationship between resistance and lateral pressure loading would allow adjustments to be made to the output reading if imperfect contact were to occur. (C,D) The previously mentioned sweep was simulated using Matlab’s FEA tool. Pressure varied between each simulation to maintain a total force of 0.4 N. The above figures display two completed simulations used for this test. The left figure represents pressure applied to the surface area less than 1.5 mm, or half of the entire sensor. The figure on the right corresponds to values less than 1.0 mm. The total area of force exerted is displayed by the black dotted outline.

• 

  Figure 10. The resistance as a function of lateral pressure loading was determined. A lateral compression defined by Figure 9 was simulated at a constant 0.4 N force. A quadratic function was fit to the to the data points, with the equation of the line displayed.