Clinical Neurophysiology 126 (2015) 1879–1885
Contents lists available at ScienceDirect
Clinical Neurophysiology journal homepage: www.elsevier.com/locate/clinph
The StartReact effect in tasks requiring end-point accuracy J.M. Castellote a,b,⇑, J. Valls-Solé c a
National School of Occupational Medicine, Carlos III Institute of Health, Madrid, Spain Department of Physical Medicine and Rehabilitation, School of Medicine, Complutense University of Madrid, Madrid, Spain c Unidad de EMG y Control Motor, Servei de Neurologia, Hospital Clínico, Universidad de Barcelona, Barcelona, Spain b
See Preface, pages 1869–1870
a r t i c l e
i n f o
Article history: Accepted 28 January 2015 Available online 18 February 2015 Keywords: Accuracy Startle Motor task Motor control
h i g h l i g h t s A startling auditory stimulus (SAS) speeds up the initial part of movement execution in tasks requiring
accuracy, as it happens with open-loop ballistic tasks. If SAS is applied once the programme has been launched, it does not interfere with its execution. The StartReact effect is restricted to movement onset, while the slow phase adjustment (homing)
depends on sensory feedback.
a b s t r a c t Objective: Fast and accurate movements are often performed in response to a sensory signal. In reaction time tasks, execution of open loop movements is speeded up when a startling auditory stimulus (SAS) is applied together with the imperative signal (IS). In this study, we examined the effects of a SAS on the performance of a task that demands accuracy. Methods: Nine subjects were asked to move a monitored pen to a target point located in a table at a fixed angular distance of 30 degrees from a start point. The target was a spot of three possible diameters: 5, 10, and 20 mm. Finger force for pen holding, pen tip pressure against the table and kinematic variables of the forearm movement were measured for three conditions: control, SAS delivered at IS (SAS-IS trials) and SAS delivered during movement execution (SAS-MOV trials). Results: Two movement phases could be identified in the movement trajectory and force profile. The first phase, ballistic, was significantly shortened in SAS-MOV trials, with earlier and larger peak velocity and peak force with respect to control trials. The second phase, slow approach to target, was longer in SAS-IS trials but not in SAS-MOV trials. Accuracy was maintained throughout all conditions and stimulation modes. Conclusions: A SAS speeds up only the first (ballistic) part of the movement in an accuracy task. Slower target approach compensates for the accelerated initial movement. No changes in the last part of the movement are seen when a SAS is delivered after movement onset. Significance: The StartReact effect is restricted to the onset of a complex movement, when muscles are activated in a ballistic mode, without feedback. Ó 2015 International Federation of Clinical Neurophysiology. Published by Elsevier Ireland Ltd. All rights reserved.
1. Introduction Some complex voluntary movements require a high level of dexterity and accuracy. For instance, object displacement is usually ⇑ Corresponding author at: National School of Occupational Medicine, Carlos III Institute of Health, C/Sinesio Delgado 6, Madrid, Spain. Tel./fax: +34 918224014. E-mail address: [email protected] (J.M. Castellote).
performed fast, with an adequate grip force to maintain the object steady and timely. When performing a skilled limb displacement, it is considered that the grip force is related to limb acceleration (Nowak and Hermsdörfer, 2004; Hermsdörfer et al., 2011; Nowak et al., 2013). This suggests that specific motor actions are part of a preconceived motor plan. However, timely execution of some corrective actions indicates also that the motor system responds to sensory inputs acting as imperative signals (ISs) along
http://dx.doi.org/10.1016/j.clinph.2015.01.028 1388-2457/Ó 2015 International Federation of Clinical Neurophysiology. Published by Elsevier Ireland Ltd. All rights reserved.
1880
J.M. Castellote, J. Valls-Solé / Clinical Neurophysiology 126 (2015) 1879–1885
movement execution. For proficient execution of simple tasks on demand, subjects may use pre-programmed movements that should be ready for execution at perception of the IS (Henderson and Dittrich, 1998). The ability to perform rapid and accurate actions would be an advantage in these cases but, according to Fitt’s law, there should be a trade-off between accuracy and speed that would prevent execution of perfectly accurate tasks in speeded up movements (Plamondon and Alimi, 1997). It is known that voluntary reactions can be speeded up by a startling auditory stimulus (SAS) delivered at the same time as the IS, a phenomenon termed StartReact (Valls-Solé et al., 1999; Carlsen et al., 2004). The same phenomenon has been described in relatively complex movements, where tasks, mainly open-loop, are temporally advanced in such a way that there is also shortening of their termination (Reynolds and Day, 2007; Queralt et al., 2010; Castellote et al., 2012). However, the phenomenon has not been studied so far in tasks in which the main requirement is accuracy and any change during task execution may have an unwanted effect. With these premises in mind, we considered the hypothesis that a SAS may disrupt the execution of a task requiring accuracy when presented (a) together with the IS, or (b) close to end-point reaching. By presenting the SAS together with the IS, we aimed at knowing whether the necessary commands for reaching end-point accuracy are packed together with the initial ballistic movement, or they are modified on-line interfering with accuracy during execution. By presenting the SAS near to end-point reaching, we aimed at knowing if the required end-point adjustments are permeable or not to external interferences. 2. Methods 2.1. Subjects Nine healthy subjects (four females and five males, aged 28– 55 years) took part in the experiment. All were self-reported right handers with normal or corrected-to-normal vision, and were free
from any neurological deficit that could affect the execution of the task. Subjects gave their informed consent for the experiment, which was approved by the local Ethics Committee. 2.2. Set up Subjects were comfortably sitting on a chair in front of a drawing table, whose surface was inclined 30°. The table had two marks: the starting point and the target. The starting point was centred at subject’s midline, 20 cm from the body and the target adjusted for each subject at 30 angular degrees to the right for a straight elbow extension movement. Both, the starting point and the target were visible to the subject at all times. Subjects were requested to hold with their right hand a home-made pen that monitored, through two strain gauge systems, the pinch grip force of the subject’s fingers during the hold, and also the force at the pen-tip during table contact (Biontec, Barcelona). The subject’s task was to move the pen from the starting point to the target. The departure point was a fixed 5 mm diameter spot. The end-point spot had three possible diameters: 5, 10 and 20 mm (Fig. 1A). An electrogoniometer (Model X 65; Biometrics; Gwent, UK) was fixed at the elbow to record forearm angular displacement reflecting pen motion, which allowed for off-line calculation of time-dependent kinematic variables. Adequate switches, one on pen tip, one on the departure spot and one on the target spot gave accurate information about departure and arrival times of the pen tip during task performance. 2.3. Procedure Subjects were told to move the pen as quickly as possible between the two targets while ensuring they landed within the second target following a somatosensory IS (a weak electrical stimulus on their left index finger). The IS was delivered after a verbal forewarning (the word ‘ready’) by pressing a computer key hidden to the subject’s vision. The time interval between forewarning and IS was variable between 1 and 3 s. Subjects
Fig. 1. Schematic representation of the set-up (A) and recordings from a representative subject (B). IS: imperative signal; RT = reaction time; MD = movement duration; RP = raising phase; BP = ballistic phase; SP = slow phase; PV = peak velocity; tPV = time to peak velocity; FD = pinch force developed; tPF = time to peak force. Tip force was used as the marker for take-off and end of the movement, while the goniometer and velocity profiles were used for determining movement phases.
J.M. Castellote, J. Valls-Solé / Clinical Neurophysiology 126 (2015) 1879–1885
were asked to perform the movement from left to right without touching the table or making a wide bow. They were allowed to practice a sufficient number of trials, with feedback from position and trajectory recordings. Stronger emphasis was placed on being accurate in placing the pen tip within the target circle, and recommendations for a more accurate movement performance were issued during practice trials. They were warned that some trials would contain a loud auditory stimulus that they should disregard (i.e., not use as any instruction in relation to task performance). They were allowed to hear the SAS once or twice but not at the same time as they were practicing the experimental task. Data collection began when subjects felt confident with their performance. No additional instructions were given. Neither verbal nor visual motivation was provided during trials. A block of 50 trials was presented for each target size. Each block was composed of 40 control trials (as described) and 10 test trials. The only difference in test trials with respect to control trials was the presence of a SAS, produced by the discharge of a magnetic coil on top of a metallic platform (Valls-Solé et al., 1999). This was made to appear in 5 trials at the time of the IS (SAS-IS trials) and in 5 other trials at the time when movement had already started, 300 ms after IS (SAS-MOV trials). Test trials were intermingled pseudorandomly with control trials in such a way that there was a minimum of three control trials between two SAS trials. The effectiveness of the SAS in inducing a startle response was checked by monitoring the EMG activity of the orbicularis oculi muscle. SAS trials with no orbicularis oculi response were excluded, and additional control and test trials were included until block completion. Interval between consecutive trials was at least 10 s. A pre-IS period of 2 s was recorded in all trials. A trial was also repeated if the pre-IS segment showed instability of force or goniometric signals. 2.4. Data storage and analysis Data were stored in a personal computer, at a sample rate of 1000 Hz, for off-line analysis. For each trial, we recorded the IS artefact, which was assigned time 0 ms for the purpose of all time calculations, the force exerted by the fingers to hold the pen, the pressure at pen tip on the table, and the displacement of the forearm with respect to time. On-line analysis of the electronic signals was done with the software package Acknowledge MP100 (Biopac Systems, Bionic Ibérica S.A., Barcelona). We considered three main events in each trial: 1. The time of pen separation from the departure point, 2. The forearm angular displacement (pen motion) and related kinematic variables of velocity and acceleration, and 3. The time of pen tip contact again with the table and the location of the point of contact with respect to the target. The actual variables measured are depicted in a representative trial in Fig. 1B, and can be divided in two groups: time variables, and kinematic and force variables. The time variables were the following: 1. Reaction time. This was measured from IS to onset of pen displacement, determined as the moment in which the pen tip lost contact with the table (take-off). 2. Movement duration. This was the time elapsed between pen tip take-off and pen contact with the table at the target area. It was measured from the recording of pen tip force (see below). The kinematic and force variables were as follows: 1. Pen trajectory. This was the recording of the goniometer signal covering the extension performed by the forearm, from pen tip take-off to pen contact within the target.
1881
2. Peak velocity. This was the peak velocity reached during the ballistic phase. 3. Time to peak velocity. This was the time from take-off to peak velocity. 4. Force developed. This was measured as the area under the curve in the pinch grip force plot, and measured from take-off to the peak force found during the ballistic phase. For each subject, values are considered in arbitrary units (a.u.) as relative percentages of those for control trials. 5. Time to peak force. This was the latency of the peak of maximum force, usually just before peak velocity in tasks requiring object displacement (Smith and Soechting, 2005). Based on the combined analysis of pen trajectory and force signals, the movement was divided in three phases (Fig. 1B): (1) Raising phase (RP), from pen tip take-off to the onset of elbow extension, (2) Ballistic phase (BP) towards target, which was the time spent in forearm extension, from the end of the raising phase to the first change in direction of the goniometer signal, and (3) Slow phase (SP) for end movement adjustment, from the end of the ballistic phase to the time of pen contact at the target area. 2.5. Statistical analysis We calculated the mean values for all variables for each subject, and obtained the great mean for each variable. Data were grouped according to two factors: Target size (5, 10, and 20 mm) and Condition (control, SAS-IS, SAS-MOV). The influence of SAS was analyzed with 2-factor repeated measures ANOVA (with Target size and Condition as within-subject factors). Statistical significance was considered at p < 0.05. Mauchly’s test was used to test sphericity, and in case of nonfulfillment, degrees of freedom were corrected by using Greenhouse–Geisser estimates of sphericity (e < 0.75) or the Huynh–Feldt correction (e > 0.75). Post-hoc pairwise comparisons were done with Bonferroni test to compare levels when the main effect of a factor was significant. For graphic representation, we chose to show data as means ± SD. 3. Results 3.1. General overview All subjects performed the task correctly after a few practice trials. They all were able to maintain a relaxed stable state, with no recognizable change in pinch or tip force recording before reacting to the IS. Also, the goniometric signal remained stable in most subjects, with only 10 trials showing a degree of instability of them during the pre-IS segment that made us repeat the trial. The mean linear distance between the starting and landing points was 22 ± 3 cm, which varied as a function of the individual’s forearm length. A few subjects failed to stop in the target area in a few trials (2.1 ± 0.2% in control trials and 3.2 ± 0.3% in test trials; X2 = 0.6, p = 0.4). We excluded test trials sporadically because of absence of startle response in the orbicularis oculi muscle, with no more than one per subject. 3.2. Movement execution profile and effects of SAS Fig. 1B shows a typical recording from one subject in a control trial with small-size target. Pen trajectory showed the three forearm movement phases described above, with an initial discrete movement to separate the pen tip from the table, a fast S-shaped ballistic displacement towards the target, and a series of small, relatively long-lasting slow oscillations until target contact. The velocity profile showed an initial negative deflection corresponding to take-off, which was followed by a large, slightly
1882
J.M. Castellote, J. Valls-Solé / Clinical Neurophysiology 126 (2015) 1879–1885
asymmetrical bell-shaped event. Pen tip force recordings showed a stable pre-IS level that went straight to 0 at movement onset and marked a sharp increase when pen tip touched again the table at the target area. Pinch force showed also pre-IS stable values, which increased after take-off. The temporal profile of the force recording during movement was slightly irregular, but a maximum peak force was observed at about the middle of the ballistic phase, synchronous with peak velocity. It was followed by a decline and small oscillations during the slow phase. A rise in force occurred also at the time of pen tip contact with the table at the target area, reflecting the stabilization of pen position.
In SAS-IS trials, subjects showed the expected advance in task onset, which led to an earlier time to peak velocity and time to peak force in SAS-IS trials in comparison to the two other trial types (Fig. 2). This was not the case for data in SAS-MOV trials, which were not different from control. Mean and standard deviation values are shown in the form of bar histograms in Figs. 3 and 4. F-value and p-value of all statistical comparisons (two-factors, repeated measures ANOVA) are summarized in Table 1 for time and kinematic variables.
3.3. Time variables
Fig. 2. Three recordings from one subject with small target for all conditions. IS: imperative signal. SAS is represented as a triangle at the time of the IS for the SAS-IS condition and 300 ms after IS for the SAS-MOV condition. Vertical dotted lines mark the presentation of the IS in all conditions. Take-off is marked with an asterisk while the time of contact is marked with a small arrow and the letter ‘c’. Note the leftwards displacement of the whole movement in SAS-IS trials but not in SAS-MOV trials.
There were significant differences in time variables due to target size and conditions (Table 1 and Fig. 3). Specific results of post hoc contrasts relevant for the aims of the study were as follows: In reaction time, the analysis of the effects of target size showed that it was shorter in large-size targets than in medium-size and small-size targets, as well as in medium-size targets with respect to small-size targets (p < 0.05 for all comparisons). The analysis of the effects of condition showed that reaction time was shorter in SAS-IS than in the other two conditions (p < 0.05). In movement duration, the analysis of target size showed significantly shorter movement duration to larger-size targets than to medium-size or small-size targets (p < 0.05) and to medium-size targets than to small-size targets (p < 0.05). The analysis of the effects of condition showed significantly shorter movement duration for SAS-IS than for the other two conditions (p < 0.05). Movement phases were influenced differently by the setup (Table 1 and Fig. 4). Mean duration and standard deviation for each movement phase and condition are reported in Table 2. All three movement phases (RP, BP and SP) were longest for small-size targets, with a significant inverse correlation between target size and duration (correlation coefficient = 0.7, p < 0.05). RP and BP were longer for control and SAS-MOV trials than for SAS-IS trials. On the contrary, SP was longer for SAS-IS trials than for the other two conditions (p < 0.001 for all comparisons).
Fig. 3. Reaction time, time to peak velocity and time to peak force for the three target sizes and conditions for all subjects. The length of the bars represent mean values while whiskers represent one standard deviation.
1883
J.M. Castellote, J. Valls-Solé / Clinical Neurophysiology 126 (2015) 1879–1885
Fig. 4. Movement duration and movement phases: raising phase, ballistic phase and slow phase for the three target sizes and conditions for all subjects. The length of the bars represent mean values while whiskers represent one standard deviation.
Table 1 Data on statistical analysis of time and kinematic/force variables. Time variables Reaction time Size Condition (SAS) Size * Condition
F
p
Kinematic variables
Table 2 Data on kinematic variables. F
p
Parameter
0.019
Contents lists available at ScienceDirect
Clinical Neurophysiology journal homepage: www.elsevier.com/locate/clinph
The StartReact effect in tasks requiring end-point accuracy J.M. Castellote a,b,⇑, J. Valls-Solé c a
National School of Occupational Medicine, Carlos III Institute of Health, Madrid, Spain Department of Physical Medicine and Rehabilitation, School of Medicine, Complutense University of Madrid, Madrid, Spain c Unidad de EMG y Control Motor, Servei de Neurologia, Hospital Clínico, Universidad de Barcelona, Barcelona, Spain b
See Preface, pages 1869–1870
a r t i c l e
i n f o
Article history: Accepted 28 January 2015 Available online 18 February 2015 Keywords: Accuracy Startle Motor task Motor control
h i g h l i g h t s A startling auditory stimulus (SAS) speeds up the initial part of movement execution in tasks requiring
accuracy, as it happens with open-loop ballistic tasks. If SAS is applied once the programme has been launched, it does not interfere with its execution. The StartReact effect is restricted to movement onset, while the slow phase adjustment (homing)
depends on sensory feedback.
a b s t r a c t Objective: Fast and accurate movements are often performed in response to a sensory signal. In reaction time tasks, execution of open loop movements is speeded up when a startling auditory stimulus (SAS) is applied together with the imperative signal (IS). In this study, we examined the effects of a SAS on the performance of a task that demands accuracy. Methods: Nine subjects were asked to move a monitored pen to a target point located in a table at a fixed angular distance of 30 degrees from a start point. The target was a spot of three possible diameters: 5, 10, and 20 mm. Finger force for pen holding, pen tip pressure against the table and kinematic variables of the forearm movement were measured for three conditions: control, SAS delivered at IS (SAS-IS trials) and SAS delivered during movement execution (SAS-MOV trials). Results: Two movement phases could be identified in the movement trajectory and force profile. The first phase, ballistic, was significantly shortened in SAS-MOV trials, with earlier and larger peak velocity and peak force with respect to control trials. The second phase, slow approach to target, was longer in SAS-IS trials but not in SAS-MOV trials. Accuracy was maintained throughout all conditions and stimulation modes. Conclusions: A SAS speeds up only the first (ballistic) part of the movement in an accuracy task. Slower target approach compensates for the accelerated initial movement. No changes in the last part of the movement are seen when a SAS is delivered after movement onset. Significance: The StartReact effect is restricted to the onset of a complex movement, when muscles are activated in a ballistic mode, without feedback. Ó 2015 International Federation of Clinical Neurophysiology. Published by Elsevier Ireland Ltd. All rights reserved.
1. Introduction Some complex voluntary movements require a high level of dexterity and accuracy. For instance, object displacement is usually ⇑ Corresponding author at: National School of Occupational Medicine, Carlos III Institute of Health, C/Sinesio Delgado 6, Madrid, Spain. Tel./fax: +34 918224014. E-mail address: [email protected] (J.M. Castellote).
performed fast, with an adequate grip force to maintain the object steady and timely. When performing a skilled limb displacement, it is considered that the grip force is related to limb acceleration (Nowak and Hermsdörfer, 2004; Hermsdörfer et al., 2011; Nowak et al., 2013). This suggests that specific motor actions are part of a preconceived motor plan. However, timely execution of some corrective actions indicates also that the motor system responds to sensory inputs acting as imperative signals (ISs) along
http://dx.doi.org/10.1016/j.clinph.2015.01.028 1388-2457/Ó 2015 International Federation of Clinical Neurophysiology. Published by Elsevier Ireland Ltd. All rights reserved.
1880
J.M. Castellote, J. Valls-Solé / Clinical Neurophysiology 126 (2015) 1879–1885
movement execution. For proficient execution of simple tasks on demand, subjects may use pre-programmed movements that should be ready for execution at perception of the IS (Henderson and Dittrich, 1998). The ability to perform rapid and accurate actions would be an advantage in these cases but, according to Fitt’s law, there should be a trade-off between accuracy and speed that would prevent execution of perfectly accurate tasks in speeded up movements (Plamondon and Alimi, 1997). It is known that voluntary reactions can be speeded up by a startling auditory stimulus (SAS) delivered at the same time as the IS, a phenomenon termed StartReact (Valls-Solé et al., 1999; Carlsen et al., 2004). The same phenomenon has been described in relatively complex movements, where tasks, mainly open-loop, are temporally advanced in such a way that there is also shortening of their termination (Reynolds and Day, 2007; Queralt et al., 2010; Castellote et al., 2012). However, the phenomenon has not been studied so far in tasks in which the main requirement is accuracy and any change during task execution may have an unwanted effect. With these premises in mind, we considered the hypothesis that a SAS may disrupt the execution of a task requiring accuracy when presented (a) together with the IS, or (b) close to end-point reaching. By presenting the SAS together with the IS, we aimed at knowing whether the necessary commands for reaching end-point accuracy are packed together with the initial ballistic movement, or they are modified on-line interfering with accuracy during execution. By presenting the SAS near to end-point reaching, we aimed at knowing if the required end-point adjustments are permeable or not to external interferences. 2. Methods 2.1. Subjects Nine healthy subjects (four females and five males, aged 28– 55 years) took part in the experiment. All were self-reported right handers with normal or corrected-to-normal vision, and were free
from any neurological deficit that could affect the execution of the task. Subjects gave their informed consent for the experiment, which was approved by the local Ethics Committee. 2.2. Set up Subjects were comfortably sitting on a chair in front of a drawing table, whose surface was inclined 30°. The table had two marks: the starting point and the target. The starting point was centred at subject’s midline, 20 cm from the body and the target adjusted for each subject at 30 angular degrees to the right for a straight elbow extension movement. Both, the starting point and the target were visible to the subject at all times. Subjects were requested to hold with their right hand a home-made pen that monitored, through two strain gauge systems, the pinch grip force of the subject’s fingers during the hold, and also the force at the pen-tip during table contact (Biontec, Barcelona). The subject’s task was to move the pen from the starting point to the target. The departure point was a fixed 5 mm diameter spot. The end-point spot had three possible diameters: 5, 10 and 20 mm (Fig. 1A). An electrogoniometer (Model X 65; Biometrics; Gwent, UK) was fixed at the elbow to record forearm angular displacement reflecting pen motion, which allowed for off-line calculation of time-dependent kinematic variables. Adequate switches, one on pen tip, one on the departure spot and one on the target spot gave accurate information about departure and arrival times of the pen tip during task performance. 2.3. Procedure Subjects were told to move the pen as quickly as possible between the two targets while ensuring they landed within the second target following a somatosensory IS (a weak electrical stimulus on their left index finger). The IS was delivered after a verbal forewarning (the word ‘ready’) by pressing a computer key hidden to the subject’s vision. The time interval between forewarning and IS was variable between 1 and 3 s. Subjects
Fig. 1. Schematic representation of the set-up (A) and recordings from a representative subject (B). IS: imperative signal; RT = reaction time; MD = movement duration; RP = raising phase; BP = ballistic phase; SP = slow phase; PV = peak velocity; tPV = time to peak velocity; FD = pinch force developed; tPF = time to peak force. Tip force was used as the marker for take-off and end of the movement, while the goniometer and velocity profiles were used for determining movement phases.
J.M. Castellote, J. Valls-Solé / Clinical Neurophysiology 126 (2015) 1879–1885
were asked to perform the movement from left to right without touching the table or making a wide bow. They were allowed to practice a sufficient number of trials, with feedback from position and trajectory recordings. Stronger emphasis was placed on being accurate in placing the pen tip within the target circle, and recommendations for a more accurate movement performance were issued during practice trials. They were warned that some trials would contain a loud auditory stimulus that they should disregard (i.e., not use as any instruction in relation to task performance). They were allowed to hear the SAS once or twice but not at the same time as they were practicing the experimental task. Data collection began when subjects felt confident with their performance. No additional instructions were given. Neither verbal nor visual motivation was provided during trials. A block of 50 trials was presented for each target size. Each block was composed of 40 control trials (as described) and 10 test trials. The only difference in test trials with respect to control trials was the presence of a SAS, produced by the discharge of a magnetic coil on top of a metallic platform (Valls-Solé et al., 1999). This was made to appear in 5 trials at the time of the IS (SAS-IS trials) and in 5 other trials at the time when movement had already started, 300 ms after IS (SAS-MOV trials). Test trials were intermingled pseudorandomly with control trials in such a way that there was a minimum of three control trials between two SAS trials. The effectiveness of the SAS in inducing a startle response was checked by monitoring the EMG activity of the orbicularis oculi muscle. SAS trials with no orbicularis oculi response were excluded, and additional control and test trials were included until block completion. Interval between consecutive trials was at least 10 s. A pre-IS period of 2 s was recorded in all trials. A trial was also repeated if the pre-IS segment showed instability of force or goniometric signals. 2.4. Data storage and analysis Data were stored in a personal computer, at a sample rate of 1000 Hz, for off-line analysis. For each trial, we recorded the IS artefact, which was assigned time 0 ms for the purpose of all time calculations, the force exerted by the fingers to hold the pen, the pressure at pen tip on the table, and the displacement of the forearm with respect to time. On-line analysis of the electronic signals was done with the software package Acknowledge MP100 (Biopac Systems, Bionic Ibérica S.A., Barcelona). We considered three main events in each trial: 1. The time of pen separation from the departure point, 2. The forearm angular displacement (pen motion) and related kinematic variables of velocity and acceleration, and 3. The time of pen tip contact again with the table and the location of the point of contact with respect to the target. The actual variables measured are depicted in a representative trial in Fig. 1B, and can be divided in two groups: time variables, and kinematic and force variables. The time variables were the following: 1. Reaction time. This was measured from IS to onset of pen displacement, determined as the moment in which the pen tip lost contact with the table (take-off). 2. Movement duration. This was the time elapsed between pen tip take-off and pen contact with the table at the target area. It was measured from the recording of pen tip force (see below). The kinematic and force variables were as follows: 1. Pen trajectory. This was the recording of the goniometer signal covering the extension performed by the forearm, from pen tip take-off to pen contact within the target.
1881
2. Peak velocity. This was the peak velocity reached during the ballistic phase. 3. Time to peak velocity. This was the time from take-off to peak velocity. 4. Force developed. This was measured as the area under the curve in the pinch grip force plot, and measured from take-off to the peak force found during the ballistic phase. For each subject, values are considered in arbitrary units (a.u.) as relative percentages of those for control trials. 5. Time to peak force. This was the latency of the peak of maximum force, usually just before peak velocity in tasks requiring object displacement (Smith and Soechting, 2005). Based on the combined analysis of pen trajectory and force signals, the movement was divided in three phases (Fig. 1B): (1) Raising phase (RP), from pen tip take-off to the onset of elbow extension, (2) Ballistic phase (BP) towards target, which was the time spent in forearm extension, from the end of the raising phase to the first change in direction of the goniometer signal, and (3) Slow phase (SP) for end movement adjustment, from the end of the ballistic phase to the time of pen contact at the target area. 2.5. Statistical analysis We calculated the mean values for all variables for each subject, and obtained the great mean for each variable. Data were grouped according to two factors: Target size (5, 10, and 20 mm) and Condition (control, SAS-IS, SAS-MOV). The influence of SAS was analyzed with 2-factor repeated measures ANOVA (with Target size and Condition as within-subject factors). Statistical significance was considered at p < 0.05. Mauchly’s test was used to test sphericity, and in case of nonfulfillment, degrees of freedom were corrected by using Greenhouse–Geisser estimates of sphericity (e < 0.75) or the Huynh–Feldt correction (e > 0.75). Post-hoc pairwise comparisons were done with Bonferroni test to compare levels when the main effect of a factor was significant. For graphic representation, we chose to show data as means ± SD. 3. Results 3.1. General overview All subjects performed the task correctly after a few practice trials. They all were able to maintain a relaxed stable state, with no recognizable change in pinch or tip force recording before reacting to the IS. Also, the goniometric signal remained stable in most subjects, with only 10 trials showing a degree of instability of them during the pre-IS segment that made us repeat the trial. The mean linear distance between the starting and landing points was 22 ± 3 cm, which varied as a function of the individual’s forearm length. A few subjects failed to stop in the target area in a few trials (2.1 ± 0.2% in control trials and 3.2 ± 0.3% in test trials; X2 = 0.6, p = 0.4). We excluded test trials sporadically because of absence of startle response in the orbicularis oculi muscle, with no more than one per subject. 3.2. Movement execution profile and effects of SAS Fig. 1B shows a typical recording from one subject in a control trial with small-size target. Pen trajectory showed the three forearm movement phases described above, with an initial discrete movement to separate the pen tip from the table, a fast S-shaped ballistic displacement towards the target, and a series of small, relatively long-lasting slow oscillations until target contact. The velocity profile showed an initial negative deflection corresponding to take-off, which was followed by a large, slightly
1882
J.M. Castellote, J. Valls-Solé / Clinical Neurophysiology 126 (2015) 1879–1885
asymmetrical bell-shaped event. Pen tip force recordings showed a stable pre-IS level that went straight to 0 at movement onset and marked a sharp increase when pen tip touched again the table at the target area. Pinch force showed also pre-IS stable values, which increased after take-off. The temporal profile of the force recording during movement was slightly irregular, but a maximum peak force was observed at about the middle of the ballistic phase, synchronous with peak velocity. It was followed by a decline and small oscillations during the slow phase. A rise in force occurred also at the time of pen tip contact with the table at the target area, reflecting the stabilization of pen position.
In SAS-IS trials, subjects showed the expected advance in task onset, which led to an earlier time to peak velocity and time to peak force in SAS-IS trials in comparison to the two other trial types (Fig. 2). This was not the case for data in SAS-MOV trials, which were not different from control. Mean and standard deviation values are shown in the form of bar histograms in Figs. 3 and 4. F-value and p-value of all statistical comparisons (two-factors, repeated measures ANOVA) are summarized in Table 1 for time and kinematic variables.
3.3. Time variables
Fig. 2. Three recordings from one subject with small target for all conditions. IS: imperative signal. SAS is represented as a triangle at the time of the IS for the SAS-IS condition and 300 ms after IS for the SAS-MOV condition. Vertical dotted lines mark the presentation of the IS in all conditions. Take-off is marked with an asterisk while the time of contact is marked with a small arrow and the letter ‘c’. Note the leftwards displacement of the whole movement in SAS-IS trials but not in SAS-MOV trials.
There were significant differences in time variables due to target size and conditions (Table 1 and Fig. 3). Specific results of post hoc contrasts relevant for the aims of the study were as follows: In reaction time, the analysis of the effects of target size showed that it was shorter in large-size targets than in medium-size and small-size targets, as well as in medium-size targets with respect to small-size targets (p < 0.05 for all comparisons). The analysis of the effects of condition showed that reaction time was shorter in SAS-IS than in the other two conditions (p < 0.05). In movement duration, the analysis of target size showed significantly shorter movement duration to larger-size targets than to medium-size or small-size targets (p < 0.05) and to medium-size targets than to small-size targets (p < 0.05). The analysis of the effects of condition showed significantly shorter movement duration for SAS-IS than for the other two conditions (p < 0.05). Movement phases were influenced differently by the setup (Table 1 and Fig. 4). Mean duration and standard deviation for each movement phase and condition are reported in Table 2. All three movement phases (RP, BP and SP) were longest for small-size targets, with a significant inverse correlation between target size and duration (correlation coefficient = 0.7, p < 0.05). RP and BP were longer for control and SAS-MOV trials than for SAS-IS trials. On the contrary, SP was longer for SAS-IS trials than for the other two conditions (p < 0.001 for all comparisons).
Fig. 3. Reaction time, time to peak velocity and time to peak force for the three target sizes and conditions for all subjects. The length of the bars represent mean values while whiskers represent one standard deviation.
1883
J.M. Castellote, J. Valls-Solé / Clinical Neurophysiology 126 (2015) 1879–1885
Fig. 4. Movement duration and movement phases: raising phase, ballistic phase and slow phase for the three target sizes and conditions for all subjects. The length of the bars represent mean values while whiskers represent one standard deviation.
Table 1 Data on statistical analysis of time and kinematic/force variables. Time variables Reaction time Size Condition (SAS) Size * Condition
F
p
Kinematic variables
Table 2 Data on kinematic variables. F
p
Parameter
0.019
