Encoding of direction of fingertip forces by human tactile afferents

Abstract

In most manipulations, we use our fingertips to apply time-varying forces to the target object in controlled directions. Here we used microneurography to assess how single tactile afferents encode the direction of fingertip forces at magnitudes, rates, and directions comparable to those arising in everyday manipulations. Using a flat stimulus surface, we applied forces to a standard site on the fingertip while recording impulse activity in 196 tactile afferents with receptive fields distributed over the entire terminal phalanx. Forces were applied in one of five directions: normal force and forces at a 20 degrees angle from the normal in the radial, distal, ulnar, or proximal directions. Nearly all afferents responded, and the responses in most slowly adapting (SA)-I, SA-II, and fast adapting (FA)-I afferents were broadly tuned to a preferred direction of force. Among afferents of each type, the preferred directions were distributed in all angular directions with reference to the stimulation site, but not uniformly. The SA-I population was biased for tangential force components in the distal direction, the SA-II population was biased in the proximal direction, and the FA-I population was biased in the proximal and radial directions. Anisotropic mechanical properties of the fingertip and the spatial relationship between the receptive field center of the afferent and the stimulus site appeared to influence the preferred direction in a manner dependent on afferent type. We conclude that tactile afferents from the whole terminal phalanx potentially contribute to the encoding of direction of fingertip forces similar to those that occur when subjects manipulate objects under natural conditions.

Fig. 1.

Electromechanical stimulator. A, The supporting frame, with one rotational and three linear stepping motors, allowed the stimulus surface to be positioned at the chosen stimulation points on the digits. B, The stimulator was built on three linear motors coupled via rods and a common rotational joint to a force/torque transducer terminating in the stimulus surface.C, Details of one of the linear motors and its position transducer.

Fig. 2.

Forces applied to the fingertip in five different directions and vectorial estimation of the preferred directions of the afferents. A, The stimulus surface was oriented parallel to the flat portion of skin at the fingertip and was advanced, under position control, to contact the skin with a force of 0.2 N. Force stimuli were superimposed on this background contact force and delivered in the normal direction and at an angle 20° to the normal with tangential components in the distal, radial, proximal, and ulnar directions as indicated by the five arrows; the normal force was always 4 N. B, Outline of a generic finger showing the stimulation point (●) and the approximate skin area (shaded) in contact with the stimulus surface for a 4 N normal force. Note the polar coordinate conventions:Ulnar (0°), Proximal(90°), Radial (180°), and Distal (270°). C, Temporal profile of the applied forces. Each stimulus consisted of a protraction phase, a plateau phase, and a retraction phase. Thesolid and superimposed dashed curvesindicate the desired and actual resultant force, respectively, exemplified for a stimulus applied in the normal direction.D, Vectorial estimation of the preferred direction of afferents illustrated on polar plots for one typical afferent of each type stimulated in eight directions 45° apart (broken lines) and corresponding plots (solid lines) for only the stimuli in the distal, proximal, ulnar, and radial directions. The polar plot, overlaid on the outline of the generic finger, consists of straight lines joining the response magnitudes in each direction of stimulation measured as number of impulses during the protraction phase; the origin of the coordinate system is at the primary site of stimulation. The dashed and solid arrows indicate the preferred direction computed as the vector sum of the responses to eight and four forces, respectively.

Fig. 3.

X-ray analysis of fingertip deformations during active and passive application of fingertip forces to a flat surface.A, The fingertip contacted a horizontal lead plate, attached to a force transducer, at an angle of ∼30°. A thin lead plate, with an attached Plexiglas platform, was glued to the fingernail. In the top panel, the subject actively applied a vertical force of either 1 or 4 N, guided by visual feedback from a moving coil voltmeter display of the force transducer output. In the bottom panel, the subject relaxed while weights placed on the Plexiglas platform passively generated a vertical contact force of 1 or 4 N. Two contours of the fingertip are superimposed. Thedashed lines show the fingertip when held in the air, and the solid lines show the skin contour when contacting the plate with a force of 4 N. B, X-ray image for a second subject at a contact force of 4 N applied actively. Thelines with arrows, perpendicular to the contact surface, indicate the distances measured for analysis. The dorsal distance was measured between the proximal edge of the upper lead plate and the upper contour of the phalangeal bone. The volar distance was measured along the same vertical linebetween the lower contour of the phalangeal bone and the top edge of the lower lead plate. C, Changes in the dorsal and volar distances caused by contact forces of 1 and 4 N applied both actively and passively. Data are from two subjects.

Fig. 4.

Distribution of the receptive fields of the afferents projected on the generic fingertip (73 SA-I, 41 SA-II, and 72 FA-I). A, The center of thecircle shows the location of the receptive field center, and the area represents the number of impulses evoked during the protraction phase, averaged across all stimuli delivered during the regular sequence. Crosses indicate the location of receptive fields of afferents that did not respond. The shaded area of skin represents an estimate of the area of contact between the stimulus surface and the fingertip at a contact force of 4 N. B, Scatter plots show the relationship between responsiveness (number of impulses during the protraction phase averaged across all stimuli delivered during the regular sequence) and the shortest (straight line) distance between the primary site of stimulation and the center of the receptive field.

Fig. 5.

Responses of SA-I afferents to forces applied in the five principal directions. A, Responses of a single SA-I afferent with the receptive field shown on the generic finger outlines at the top left. Impulse ensembles show responses to the stimuli repeated during the regular sequence (n = 5), and histograms show the instantaneous frequency averaged over the five trials.Solid and broken lines show the forces and positions of the stimulus, respectively, averaged over the five trials. B, The generic finger outlineshows a polar plot for the afferent illustrated in A. The polar plot consists of four straight lines joining the response magnitudes in the distal, radial, proximal, and ulnar directions measured as number of impulses during the protraction phase; the origin of the coordinate system is at the primary site of stimulation. C, Overlaid polar plots, superimposed on the generic finger, for the 27 afferents for which response was greatest when the tangential component of force was in the distal direction. D, Instantaneous firing rates, averaged over the five trials, for the same 27 afferents as in C, shown for forces with tangential components in the four directions and for normal force stimulation. E, For each of the 27 afferents in C, lines join three data points representing the response, averaged over the five trials, to forces in the proximal (P), normal (N), and distal (D) directions. Broken and solid lines show afferents for which tangential force components showed a net effect determined by the Mann–Whitney U test (n = 9) and those for which there was no significant net effect (n = 18), respectively.

Fig. 6.

Preferred directions of the SA-I afferents estimated from the responses evoked during the protraction phases of the four stimuli with tangential force components. A,Arrows (unit vectors) show the preferred directions of the 68 directionally sensitive SA-I afferents. The mean of the 68 vectors is shown by the white arrow. Solid circles on contours of the generic fingertip indicate the location of the receptive field centers of afferents showing greatest responses in the distal, radial, proximal, and ulnar quadrants. Thefine dots (same on each finger) indicate the receptive field centers of all the directionally sensitive SA-I afferents.B, The overall responsiveness of the afferents (the average number of impulses evoked during the protraction phase of the four stimuli with tangential force components) as a function of the preferred direction of the afferent.

Fig. 7.

Directional sensitivity of the same 68 SA-I afferents as displayed in Figure 6. A, Directional sensitivity vectors superimposed on the generic finger.B, Magnitude of the vector, the directional sensitivity index, as a function of the preferred direction of the afferent.C, Directional sensitivity of afferents as a function of the responsiveness of the afferent, measured as the number of impulses evoked during the protraction phase. D, Relationship between the distance from the stimulation site to the receptive field center and the directional sensitivity of the afferent.

Fig. 8.

Responses of SA-II afferents to forces applied in the five principal directions. A, Responses of a single SA-II afferent. B, The generic finger outline shows a polar plot for the afferent illustrated inA. C–E, Data from 15 afferents for which response was greatest when the tangential component of force was in the proximal direction. For explanation, see legend to Figure 5.

Fig. 9.

Preferred directions of the 32 directionally sensitive SA-II afferents. For explanation, see legend to Figure6.

Fig. 10.

Directional sensitivity of the same 32 SA-II afferents displayed in Figure 9. For explanation, see legend to Figure7.

Fig. 11.

Responses of FA-I afferents to forces applied in the five principal directions. A, Responses of a single FA-I afferent with the receptive field shown on the generic finger outlines at the top left.B, The generic finger outline shows a polar plot for the afferent illustrated in A.C, Overlaid polar plots of the protraction phase, superimposed on the generic finger outline, for the 21 afferents for which the response was greatest when the tangential component of force was in the proximal direction. D, Instantaneous firing rates during the protraction phase (left), averaged over the five trials, for the same 21 afferents as in C, shown for forces with tangential components in the four directions and for normal force stimulation. On the right, the average instantaneous firing rates during the retraction phase are shown for a different sample of 16 FA-I afferents that responded best to stimuli in the proximal direction during the retraction phase. The sample is different from the sample on the left because FA-I afferents had different directional preferences during the protraction and retraction phases.E, For each of the 21 afferents illustrated inC, lines join three data points representing the response, averaged over the five trials, to forces in the proximal (P), normal (N), and distal (D) directions. For further explanation, see legend to Figure 5.

Fig. 12.

Preferred directions of the 50 directionally sensitive FA-I afferents. For explanation, see legend to Figure6.

Fig. 13.

Directional sensitivity of the same 50 FA-I afferents displayed in Figure 12. For explanation, see legend to Figure7.

Fig. 14.

Comparison of directionality during different phases of the stimulus. A, Open bars show the number of afferents that responded during that phase of the stimulus. Solid bars show the number of afferents that were directionally sensitive. Dotted lines indicate the total number of afferents of each type recorded from. B, Scatter plots displaying the relationship between the preferred direction during the protraction phase and that during the plateau phase for 61 SA-I and 31 SA-II afferents, and during the protraction and retraction phases for 29 FA-I afferents. Data are from afferents that were directionally sensitive during both phases of the regular sequence. Data outside the thin solid lines represent afferents for which the preferred directions for the two phases of stimulation differed by >90°. Data on the dotted lines would correspond to a 180° difference, which represents a reversal of the preferred direction.

Fig. 15.

Comparison of preferred directions calculated for the regular sequence with those calculated for the irregular sequence. Data points show afferents (67 SA-I, 30 SA-II, and 45 FA-I) that had a statistically significant preferred direction for both modes of stimulus presentation.

Fig. 16.

Displacement of the stimulus surface resulting from the four tangential force components during the regular sequence.A, Bars show the mean magnitude of the total displacement of the stimulus surface when forces were applied with tangential components in the ulnar (U), proximal (P), radial (R), and distal (D) directions. Vertical lines indicate 1 SD. B, Unit vectors for all afferents (n = 186) show the estimated direction of force that resulted in maximum displacement of the stimulus. Thewhite arrow shows the mean vector. C, Filled histograms show distributions of the preferred directions of the afferents, and open histograms show distributions of the directions of force producing maximum displacement in the tangential plane for the same afferents (73 SA-I, 41 SA-II, and 72 FA-I afferents).D, Angular differences between the preferred direction and the direction producing maximum displacement for data shown inC.