Smarttouch® SF (Johnson&Johnson, New
Brunswick, NJ, USA):
The tip of the catheter is suspended by a machined precision spring. A
magnetic transducer generates a small magnetic field in the distal part
of the catheter. Three magnetic field sensors are located in the more
proximal part of the catheter, distributed around the circumference. As
forces are acting on the catheter tip, it moves slightly towards the
sensors, changing the signal received by the sensor coils. By Hooke’s
law, the acting force and direction can be calculated from the three
sensors and the known characteristics of the spring.
Measurement protocol
All catheters were fixed 18-20mm proximal to the tip to get an
adequately low bending of the distal part while not compromising the
force sensor tip by the clamping mechanism. Prior to the measurements,
the catheters were submerged in the heated saline bath for five minutes,
allowing for warm-up. The catheter was zeroed after each change of the
contact angle. Measurements were taken repeatedly adjusting the exerted
force in-between measurements until a minimum of 100 measurements at one
specific angle was reached. During acquisition, equal distribution of
the measurements across the full measurement range of 0g to 60g was
ensured. Subsequently, the catheter was fixed at a different angle and
the same protocol was repeated for all angles of 0°, 30°, 45°, 60°, and
90°. The error made by the catheter was calculated by subtracting the
weight displayed on the scale (real force) from the contact force
displayed on the catheter readout (measured force). A resulting negative
value means that the catheter underestimates the real force; a positive
value means the catheter overestimates the real force. We evaluated
three catheters for each model to verify reproducibility and to quantify
inter-catheter variability. For the Stablepoint™catheter, measured-force values above 50g are not displayed at angles
>45°.
Statistical analysis
Continuous variables are presented as mean ± standard deviation or as
median and interquartile range as appropriate. Linear regression
analysis was used to determine the measurement error due to friction.
Spearman correlation coefficients were used to assess the correlation
between measured force and real force. A local smoothing function
(locally estimated scatterplot smoothing (LOESS)) was used for the
interpretation of the measurement data (loess function, span = 0.3).
Statistical analyses were made by using R 4.0.2 (R Core Team, Vienna,
Austria).