Fig 1.
Experimental setting used for studies of local palpation behavior.
Coordinate frame reflects lateral forces (Fx and Fy) and normal force (Fz) applied by subject’s index finger.
Fig 2.
Histogram of standard deviations across all trials for normalized measurements.
a) Lateral force Fx, b) lateral force Fy, and c) normal force Fz.
Fig 3.
Histogram of correlation coefficients for measurements of two connected lateral forces Fx and Fy.
Fig 4.
Constraints of the finger motion during single point palpation.
For Fx (direction is chosen for this example), the finger can move freely in both directions, but in the case of Fy (direction is chosen for this example), the motion is limited by a finger joint.
Fig 5.
Distribution of lateral force with fitted ellipses (red dashed line) for one selected subject, one trial, for different nodules a) to g): Empty, 3, 6, 9, 12, 15 and 18 mm, respectively.
Fig 6.
Distribution of frequency for the modulation of lateral forces for different probing locations for all subjects.
Fig 7.
Sample profiles of normal forces for sinusoidal (red dotted lines) and step—like response (black solid lines) from diffident subjects.
Table 1.
Selection of Order Number for Autoregressive Model.
Fig 8.
Histograms of prediction horizon and costs of NRMSE for second order reactive model, 100% assumes perfect fit.
Fig 9.
Experimental setup to validate autonomous palpation based on force modulation strategy.
Fig 10.
Example stiffness measurement for autonomous palpation.
Analysis of steady state response to calculate the average time required to perform autonomous palpation.
Fig 11.
Stiffness measurements for indentation based measurements (red dotted line) and autonomous palpation based on AR model (solid line) for silicone with no nodule, nodules of 3, 6, 9, 12 and 15 mm for all trials.
Fig 12.
a) Variance of stiffness measurements for autonomous palpation, and b) Difference of stiffness for autonomous palpation and indentation based measurement.
For silicone with no nodule, nodules of 3, 6, 9, 12 and 15 mm. Error bars show standard deviation for multiple trials.
Fig 13.
Stiffness obtained during validation experiments for nodules with 3 mm (a) and b)) and 9 mm diameter (c) and d)).
Subfigures (a) and c)) show stiffness obtained using lateral force modulation of the robotic probe only. Subfigures (b) and d)) display stiffness for palpation using normal force generated by AR based palpation with no modulation of lateral force. Red dotted line shows stiffness measurements for indentation based palpation.
Fig 14.
Results of validation studies.
a) Variance of stiffness measurements and b) Difference of stiffness for autonomous palpation and indentation based measurement, for silicone with nodules of 3 mm and 9 mm. Results from validation studies for separate lateral force modulations (Lat.) and for AR based palpation with no lateral movement. Error bars show standard deviation for multiple trials.
Fig 15.
Variance and difference of stiffness of hard nodules and soft environment for robotic palpation, for three different strategies.
Red points show the results of combined autonomous palpation, blue points correspond to validation of separate lateral motion, and black points to separate normal motion.
Table 2.
Relevant work on human palpation studies for robotic applications.