Here I present more details about the methods of the experiments which are. described in the main text, and describe two additional examinations which

Transcription

1 Supplementary Note Here I present more details about the methods of the experiments which are described in the main text, and describe two additional examinations which assessed DF s proprioceptive performance and tested her for signs of simultanagnosia. I. Methods of the experiments presented in the main text Experimental Setup The setup was designed to ensure that the computer-generated stimuli appeared in the same plane within which subjects performed their movements. To achieve this the computer-generated image is first projected (using an Epson EMP74/54 multimedia projector) onto a horizontally suspended backprojecting screen (distance between projector and horizontal screen: 61 cm), the image is then reflected on the mirror below (distance between the mirror and the screen: 33 cm), and thus appears to be on a horizontal surface which is approximately 1 cm above the table top (distance between the mirror and the table top: 33 cm; distance between subjects eyes and the centre of the image plane: 57 cm). This surface coincides with the plane as defined by the Phantom Haptic Interface (Phantom Premium 1.5, SensAble, Cambridge, MA, USA). The Phantom is a lightweight robot (footprint: 25 x 33 cm; workspace: width x height x depth: 19.5 x 27 x 37.5 cm; maximum exertable force: 8.5 N). with one arm, which consists of three segments with three joints. A thimble is attached to the tip of the arm. Subjects put their finger (in our case the index finger of their right hand) in the 1

2 thimble. The Phantom acts as force-feedback device, which can be used to generate haptic feedback. The Phantom samples the 3D position of the tip of its arm with a temporal resolution of 1kHz and spatial resolution of 0.03 mm. For the purpose of our experiment, the positional data was sampled with a frequency of 100 Hz. We used the force-feedback capacity of the Phantom to generate a virtual table-top which was suspended 1 cm above the real table top, and defined the movement plane for the subjects. This movement plane coincided with the image plane. By default the subjects were not able to see their hand. However, we could present a dot which indicated to the subject the current position of their finger tip within the movement plane. This was used to guide the subjects finger onto the start position at the beginning of the trial. During the trial visual information about the finger position was not available. Stimulus material and experimental procedures. The target dots had a diameter of 0.5 of visual angle and were either red or yellow; the height/width of the reference cross (black) used for the start position was 1.5 of visual angle. The target stimulus configuration was presented for 2 seconds. In the perceptual tasks, subjects had to indicate verbally which of the two target dots were closer to the reference position. Six combinations of distances were used (i.e. 70/77 mm; 70/84 mm; 70/98 mm; 70/112 mm; 70/140 mm; 70/170 mm). Twenty trials per combination were presented. In half of the trials the smaller distance was to the right. The reference position changed from trial to trial. Trials were presented in a random order. In the egocentric version of the perceptual task the reference position was defined by the tip of the subject s right index finger. The index finger 2

3 was placed in the thimble at a specific position within the movement plane. A small cross indicated the position for the finger at the beginning of each trial. Subjects received visual feedback to allow them to guide their finger to the predetermined finger position. Both the cross and the cursor indicating the current finger position were turned off once the finger had arrived at the indicated position. The same procedure was used to guide subjects fingers to the start position in the visuomotor task. Only one target presented at a variable horizontal distance from the reference cross (allocentric condition) or the start position (egocentric condition) was used in the visuomotor task. Five different target distances were used (46, 55, 74, 102 and 139 mm). Target distance and start position varied randomly from trial to trial. Twenty trials per target distance were presented. Subjects pointing movements were recorded with the Phantom Haptic Interface. Data analysis. In the case of the two perceptual experiments, we had six different combinations of distances ranging from difficult (70 versus 77 mm; difference 7 mm). to easy (70 versus 170 mm; difference: 100 mm). The number of correct judgements for each combination were counted. The difference in distances which produced 80% or more correct judgements was then taken as the value for the distance-discrimination threshold. 3

4 The same accuracy measure was used to characterise the performance in the visuomotor tasks. This is important because it has been argued that perceptual tasks often require a higher level of visual accuracy than visuomotor task 1. Accordingly impaired perceptual performance and preserved visuomotor performance may just reflect differences in task demands and not differences in the actual accuracy with which visual information is used in the two tasks. Using the same measure to characterise accuracy in both tasks allows us to make a direct comparison between the two tasks and ensures that any differences obtained reflect genuine differences in visual accuracy. The following procedure was used to calculate the distance-discrimination threshold on the basis of the obtained visuomotor responses. For this purpose the set of amplitudes of all pointing movements obtained in trials with one specific target distance is paired with a second set of pointing amplitudes for a different target distance. Using the mean and standard deviations of the amplitude of the two sets of motor responses we calculate d as the difference between the z-transforms of the two means (see 2, pp. 118, 119). We used a table for a two-alternative forced-choice task (see 2, Appendix, Table B) to translate the d values into percentage-correct values. This was done for all combinations of target distances, and allowed us to plot the percentage of correct discrimination judgements as a function of the differences between the various combinations of distances. Finally, we determined for each condition and each 4

5 subject the 80% distance-discrimination threshold (i.e. the difference in target distance at which 80% or more correct judgements were achieved). II. Proprioceptive performance of patient DF and healthy controls We used two tasks to compare DF s proprioceptive performance with that of 10 right-hand dominant, female, age-matched (mean age: 51 yrs; range: yrs.) controls. Subjects were blindfolded for both tasks. In the first task, the experimenter positioned the subject s left index finger at some distance of their left thumb. The subject was then asked to match that distance with the index finger and the thumb of their right hand (condition: LHgrasp). Five distances were used: 3, 5, 7, 8 and 9 cm. Each distance was repeated four times yielding a total of 20 trials. The same procedure was then repeated using the right hand as the reference hand (condition: RH-grasp). In the second task, subjects had to match the position of the index finger of their left hand with the position of their right index finger. A long wooden bar, which was fixed to the table and extended horizontally away from the subject s midline, was used to ensure that the two fingers would not touch during the matching task. The left index finger was positioned along the left side of that bar at a given distance from the subject s body. The subject had to move their right index to the corresponding distance on the right side of the wooden bar (condition: LH-reach). Seven different distances (10, 15,20,25,30,35,40 cm) were used. Each distance was repeated three times, resulting in a total of 21 trials. The task was then repeated with the right hand as the reference hand (condition: RH-reach). 5

6 The finger positions were recorded using a Zebris 3D position-registration device (Fa. Zebris, Tuebingen, Germany). For task 1 we used the distance between the index and thumb; for task 2 we used the distance between the near end of the wooden bar and the position of the index finger. These distances were measured for both the estimated and reference position. For both tasks the error was calculated by computing the mean of the absolute difference between the reference distance and the estimated distance. To test whether DF s performance was in the normal range we used a modified t-test, which had been developed to compare the performance of a single patient with the results obtained for a group of control subjects 3. DF s errors were within the limits of normal performance (Supplementary Table 1). Supplementary Table 1. Proprioceptive performance of patient DF and controls Error [mm] Modified t-test Condition DF Normmean Normsd Normmin Normmax T p LH-grasp RH-grasp LH-reach RH-reach III. Testing for simultanagnosia in patient DF To test for signs of simultanagnosia in patient DF I used the same stimuli (red and yellow dots, black cross), setup and presentation time (2 seconds) as in the allocentric, perceptual task. In each trial one, two or three visual items were 6

7 displayed. Sixty trials in total (20 trials per conditions) were presented in a random sequence. DF reported the number of items without errors. In a second task, three visual items (black cross, yellow and red dot) were displayed in each trial. The black cross could be in the middle, to the left or right of the yellow and red dot. For each configuration 20 trials were generated, resulting in a total of 60 trials. Each configuration was presented for 2 seconds. DF reported the relative positions of the black cross (i.e., right, left or middle) without errors. Thus, at least for the stimulus configurations used in the present study, DF did not show signs of simultanagnosia and could indicate at least in qualitative terms (left/right/middle) the position of one visual item relative to the position of one or more other visual items. References 1. Vecera, S. P. Neuropsychologia 40, (2002). 2. Gescheider, G. A. Psychophysics: The Fundamentals. (Lawrence Erlbaum Associates, Mahwah, NJ, 1997). 3. Crawford, J.R. & Howell, D.C. Clinical Neuropsychol.12, (1998). 7