Effect of Gloving on Perceptual and Manipulation Task Performance

This study evaluates the effect of fingertip covering on the performance of perceptual and manipulation tasks. For the perceptual task, subjects were timed as they detected hard lumps in soft rubber models while barehanded and while wearing gloves of thickness 0.32 mm, 0.64 mm, 0.95 mm, 1.27 mm, 1.59 mm and 1.91 mm. Four lump sizes with diameters 3.2 mm, 4.8 mm, 6.4 mm and 7.9 mm were used. Analysis of the data yielded significant differences in lump detection time with glove thickness. Detection time variation was greatest for the 3.2 mm lump. Mean times were always best with bare hands and poorest with 1.91 mm glove thickness. The maximum force applied during palpation increased linearly with glove thickness. In the manipulation task, seven subjects were asked to lift a 460g object using the thumb and index finger while barehanded and wearing gloves of thickness 0.16 mm, 0.32 mm, 0.95 mm and 1.91 mm. The object was covered with three different surfaces with varying frictional conditions: sandpaper, suede and rayon. As glove thickness increased, the subjects’ ability to adapt to new surfaces decreased and increasing levels of excess grip force were applied. Visual feedback did not play an important role in assisting lift for any glove thickness. The results of the perceptual and manipulation tasks suggest that the effects of gloving are both thickness dependent and highly task sensitive. INTRODUCTION Distributed sensations across the fingerpad are useful in the execution of many tasks. Certain applications, however, involve conditions where cutaneous information is either inhibited or non-existent. An ubiquitous example is the wearing of gloves in medical procedures or hazardous environments. Gloves have many relevant properties including thickness, stiffness and fit; the thicker the glove, for example, the less cutaneous information passes through. The question of how gloves with different properties impair task performance therefore becomes a relevant issue. This study analyzed glove thickness as a parameter. Previous studies on the effects of gloving have indeed indicated that performance of certain tasks is impaired. A study on chemical gloves (Bensel 1993) found that wearing gloves of up to 0.64 mm thickness yielded significant differences in task performance time compared to bare hands. Tasks were chosen to test manual dexterity, such as rifle assembly and tool usage. Bensel also found that performance in these manual dexterity tasks could be improved significantly with practice. Other studies have examined the effect of double gloving in surgery as a protective measure against infection and have found that two thin gloves have negligible effect on surgical technique (Webb and Pentlow 1993, Burke et al. 1989). Subjective studies have shown, however, that double gloving is uncomfortable and undesirable to the surgeon (Wilson et al. 1996) and that there is a high negative correlation between glove stiffness and subjects’ assessment of tactile perception (Burke et al. 1989). A limitation of these earlier works is that they only tested the effects of relatively thin gloves. The goal of the present study was therefore to quantitatively examine the effects of gloving for a much wider range of thicknesses. This was motivated by the hypothesis that beyond a certain thickness, glove interference may exceed a task-impairment threshold and lead to failure of effective manipulation. Tasks were chosen from the principal types of fingertip activities: a perceptual task (lump detection) and a manipulation task (object lifting using the precision grip). Variations of these tasks have been studied in the past, although none of them have tested glove thickness as a variable. Lederman and Klatsky (1997) studied lump detection for both the barehanded case and fingers sheathed in a rigid fiberglass covering. The goal of this work was to assess the importance of incorporating spatially distributed fingertip forces in the design of haptic interfaces. These rigid coverings drastically impaired task performance. Bloom et al. (1982) had subjects palpate silicone breast models containing lumps with different size, depth, hardness and fixation properties. The goal was to determine the relationship between optimum palpation skill and tumor characteristics in order to enhance breast self-examination technique. They found that for fixed lumps, size was the most important factor in detection ability whereas hardness and depth had little effect. For mobile lumps, both size and hardness contributed significantly to detection. In light of Bloom et al.’s findings, this present study focused on fixed lumps and used size as a variable. Performance was evaluated in terms of the time required to find lumps and the maximum force applied. Use of the precision grip in object lifting was researched extensively by Johansson and Westling (1984) for the barehanded case. They quantitatively examined the coordination of grip force and vertical lifting force and found that grip force was regulated to prevent slips based on both the weight of the lifted object and the frictional condition between the object and skin. In studies that investigated the underlying neural mechanisms, Johansson and Westling (1984, 1987) determined the importance of cutaneous feedback in slip prevention. The fingers of a thick latex glove (Microflex Corp., UltraOne®, k≈90 N/m) were used for the fingertip covering. Thickness was achieved by layering different-sized glove fingers on top of each other. The full range was from zero thickness (bare hands) to a maximum of 1.91 mm (six gloves). Plaster molds of the gloves were used initially to determine the precise size of the fingers to ensure close glove fit without stretching as the layers were increased. For this study, it was hypothesized that gloving would interfere with slip sensation because it occurs at the surface of the fingertips. To test the hypothesis, this study varied the frictional condition of the object to be lifted, and measured the excessive grip force applied. A force sensor was placed under the petri dish to measure normal forces applied during palpation. Calibration of the force showed that maximum error was less than about 8%. Data was recorded at 50 Hz. The first non-zero force signal was set as t=0. METHODS Procedure At the beginning of each study, subjects were instructed to palpate the rubber models and press the signal button when a lump was found. They were informed that each tissue sample contained either three equal-sized lumps or no lumps at all. To enable unbiased testing of all lump sizes at all glove thicknesses, both were presented in a pseudo-random order, i.e. no two consecutive tasks used the same lump size nor the same glove thickness. For each test, the normal forces applied during palpation and the signals from the detection button were recorded. A maximum time limit of 180 seconds was set for each test. A. Lump detection