Content area
Traditional research approaches to the reach-to-grasp movement have employed real-world perturbations involving physical objects. Recent technological advances provide new avenues for the investigation of sensorimotor control including the use of Virtual Reality Environments (VE). In this study, we used an immersive VE to produce compelling perturbations of target object size and position and Transcranial Magnetic Stimulation (TMS) to probe the neural bases of compensatory responses during grasping movements. Extensive research has identified a Dorsolateral (DL) and a Dorsomedial (DM) pathway as the likely neural bases for the sensorimotor coordination underlying grasping movements. In order to test the causal involvement of the parietal and premotor nodes of both pathways, we implemented visual perturbations of object size and distance at two different latencies (100 and 300 ms after movement onset) with concurrent TMS in a fully randomized design. The kinematic profiles of the grasping movements exhibited clear effects of the visual perturbations, particularly the late ones. We found that TMS stimulation of aIPS during the late perturbation of object size modified the timing of aperture closing. Similarly, TMS to PMv during the late perturbation of object distance reduced transport velocity during the compensatory double-peak. Our results support the involvement of the DL pathway when quick modifications including complex digit control are required. Against our expectations, sudden changes in target position did not elicit activity in the DM pathway. This study supports the notion that VE can be successfully employed for the study of the neural substrates of motor control.
1 Introduction
Research on the reach-to-grasp movement has traditionally employed real-world perturbations involving physical objects. Nonetheless, recent technological advances have created new possibilities for the investigation of sensorimotor control, including the use of Virtual Reality (VR) environments ( Furmanek et al., 2019). The immersive nature of modern VR systems ensures that participants respond to virtual objects with movement patterns that closely approximate those observed in real-world grasping tasks, validating the ecological relevance of VR-based paradigms ( Furmanek et al., 2019). This new approach allows researchers to produce and control the timing and visual properties of target stimuli in a compelling manner while, at the same time, probing the underlying neural mechanisms through cortical stimulation such as Transcranial Magnetic Stimulation (TMS). Despite these methodological advantages, the application of TMS within VR environments for studying grasping movements remains largely unexplored. In this study, we employed an immersive VR environment to produce visual perturbations of object size and position while concurrently using TMS to target the main nodes of the frontoparietal networks controlling the reach-to-grasp.
Extensive evidence supports the notion that reach-to-grasp movements are mainly controlled by two frontoparietal networks, each composed of a parietal and a premotor (frontal) region. More specifically, the Dorsolateral (DL) pathway, consisting of the anterior intraparietal sulcus (aIPS - monkey AIP) and the ventral premotor cortex (PMv - monkey F5), ( Gerbella et al., 2017; Gharbawie et al., 2011) and the Dorsomedial (DM) pathway, consisting of the anterior superior parieto-occipital cortex (aSPOC - monkey V6A) and its connections to dorsal premotor cortex (PMd - monkey F2), ( Cavina-Pratesi et al., 2010; Fattori et al., 2017; Gamberini et al., 2009; Pitzalis et al., 2015; Raos et al., 2004).
Convergent evidence from studies in human and non-human primates has shown that reaching and grasping movements activate both the DM pathway ( Fattori et al., 2001; P. 2010; Vesia et al., 2017) and the DL pathway ( Binkofski et al., 1999; Lehmann and Scherberger, 2013; Murata et al., 2000). But, importantly, their differential connectivity results in distinct functional properties. For example, visual processing in the DM pathway appears to be more closely attuned to the execution of quick reach-to-grasp actions through the processing of veridical target position, arm transport parameters and wrist orientation ( Galletti and Fattori, 2018). In contrast, the DL pathway has been implicated in the processing of the intrinsic characteristics of the target object, such as shape and size, as well as in the fine control of digit movements ( Binkofski et al., 1998; Frey et al., 2005). Overall, the accumulated evidence suggests a functional differentiation due to sensory context and accuracy constraints.
TMS has been employed to confirm the functional relationship between the nodes in these networks gleaned from correlational work. Tunik, Frey and Grafton ( Tunik et al., 2005) pioneered this approach by demonstrating the involvement of aIPS in the flexible response to a perturbation of target size during a grasp. Subsequent research has probed and expanded our understanding of reach-to-grasp processing in the frontoparietal grasping networks ( Breveglieri et al., 2023b; Breveglieri et al., 2021; Cohen et al., 2009; Davare et al., 2007; 2010; Furmanek et al., 2025; Schettino et al., 2015; Vesia et al., 2017).
In this study, we employed TMS to disrupt processing in each of the nodes of the DM (aSPOC, PMd) and DL (aIPS, PMv) pathways and to compare their causal involvement in the correction to reach-to-grasp movements to target object size or position perturbations in a virtual reality (VR) environment. The use of a VR environment allowed us to produce compelling visual perturbations that elicited canonical kinematic modifications at two different timings during the movement (100 msec and 300 msec after movement onset), ( Furmanek et al., 2022).
We reasoned that a perturbation of object size, an intrinsic characteristic of the target, would require relatively more visual analysis and digit control to correct. Contrarily, changes in object position would require a fast computation of distance, akin to when an object moves away. Therefore, we expected to see differences in the involvement of the two pathways: the nodes of the DL pathway should be more strongly involved in coordination of the reach and grasp when correcting for object size perturbations. On the other hand, the DM pathway should be involved in implementing the response to object position perturbations ( Breveglieri et al., 2023a).
Furthermore, to examine the temporal involvement of each of the regions under study, we delivered TMS at either 100 msec ( Early) or 300 msec ( Late) after movement onset, concurrent with the visual perturbations. These latencies were chosen to capture different phases of the grasping movement, namely, wrist acceleration and time to peak velocity, respectively. The rationale for selecting these timepoints was to look at the involvement of the frontoparietal networks nodes in the corrective response during two phases of the movement where the need for sensory input is thought to differ, with later phases involving relatively more sensory feedback processing.
2 Material & method
2.1 Participants
The experiment adhered to the Declaration of Helsinki and received approval from Northeastern University's Institutional Review Board. Nine right-handed ( Oldfield, 1971; Steenhuis and Bryden, 1989) participants (mean ±SD age: 25 ± 7.3 years; 5F, 4M) with no reported muscular, orthopedic, or neurological conditions voluntarily participated after providing written and verbal informed consent.
2.2 Experimental setup
These methods were described in our previous paper ( Furmanek et al., 2025). Here, we provide brief descriptions of the experimental setup, procedure and perturbation schedule. Participants performed reach-to-grasp movements in an immersive virtual environment (VE) developed in Unity (v5.6.1f1, 64-bit, Unity Technologies Inc., San Francisco, CA) and displayed via an HTC Vive Pro head-mounted display - HMD (HTC Inc., Conway, SC). The HMD’s interpupillary distance was adjusted individually, and head motion was tracked using an Inertial Measurement Unit and embedded laser-based infrared markers. An eight-camera motion tracking system (PPT Studio NTM, WorldViz Inc., Santa Barbara, CA; sampling rate, 90 Hz) recorded 3D motion from IRED markers on the wrist, thumb, and index finger. In VE, participants viewed their thumb and index fingertips as 3D green spheres (0.8 cm diameter) ( Furmanek et al., 2021). Custom C# software-controlled trial scheduling, object rendering, and perturbation timing. Prior research confirms that reach-to-grasp coordination remains consistent for virtual and physical objects in this haptic-free VE ( Furmanek et al., 2019, 2021; Mangalam et al., 2021).
2.3 Experimental procedure and design
Seated at a table (
Fig. 1
2.4 TMS-induced cortical perturbations and visual perturbation timing
Each participant completed one session, testing all brain sites (aSPOC, PMd, aIPS, PMv) in a randomized order. Visual perturbations (VP) occurred either Early (100 ms) or Late (300 ms) after movement onset. In the Size Perturbation ( Fig. 1B), the target object (3.6 × 2.5 × 8.0 cm, width, depth, height) was positioned 24 cm from the hand. During perturbation trials, it expanded to (7.2 × 2.5 × 8.0 cm) while remaining in place. In the Distance Perturbation ( Fig. 1C), the target object (3.6 × 2.5 × 8.0 cm) started 24 cm away. In perturbation trials, it instantly shifted to a farther location (36 cm).
Each session included a total of 704 trials over ∼2.5 h, divided into four 30-minute mini-blocks. Participants took seven breaks (5 min), one every 88 trials. Each mini-block contained 176 trials, with VP in 25 % of trials. TMS was applied in 50 % of all trials, coinciding with VP. Of the 96 perturbed trials per mini-block, early and late perturbations were equally split. Each combination of perturbation timing and TMS condition (2 timings × 2 TMS conditions) occurred in 8 trials. Trial order was randomized and repeated four times per mini-block. Based on prior work (
Furmanek et al., 2019;
Schettino et al., 2015), the timing of early perturbations occurred before, and late perturbations shortly after, peak transport velocity in unrestrained movements. Control trials included 40 unperturbed trials without TMS (SNX) and 20 Early and 20 Late unperturbed trials with TMS (SNET, SNLT), see
Table 1
2.5 Neuronavigated transcranial magnetic stimulation
Each participant underwent high-resolution anatomical MRI (Siemens MAGNETOM Prisma 3T; T1-weighted 3D-MPRAGE, 1 × 1 × 1 mm voxels). MRI fiducials were co-registered with the participant’s head for frameless neuronavigation (Brainsight, Rogue Research Inc., Montreal, Canada), ensuring precise TMS targeting. TMS was delivered using a Magstim Bistim 2 with a D70 2 figure-of-eight coil (Magstim Inc., Whitland, UK). The cortical hotspot for the first dorsal interosseous (FDI) muscle was identified via coarse mapping of the left precentral gyrus hand knob ( Weiss et al., 2013). The hotspot, producing the largest motor evoked potential (MEP), was determined by peak-to-peak EMG amplitude (10–40 ms post-TMS). Resting motor threshold (RMT) was defined as the stimulator intensity ( % MSO) required to elicit MEPs >50 µV in 5/10 trials ( Rossini et al., 2015), averaging 41.1 ± 6.6 % across subjects.
2.6 Localization of brain sites and TMS
Each of the four brain sites (aSPOC, PMd, aIPS, PMv) was marked based on anatomical landmarks identified on each participant’s reconstructed volumetric MR image ( Fig. 1D), guided by established localizers from prior published fMRI and TMS studies on related tasks: (1) anterior superior parietal-occipital cortex (aSPOC), along the medial surface of the parietal lobe and just anterior / rostral to the parietal occipital (PO) sulcus, commonly referred to as the precuneate region. The aSPOC also includes the parietal convexity adjacent to and just off the midline precuneate region but medial to the intraparietal sulcus ( Cavina-Pratesi et al., 2010; Vesia et al., 2010, 2013); (2) dorsal premotor cortex (PMd), superior portion of the precentral gyrus, as delimited by the superior frontal sulcus ( Davare et al., 2006); (3) anterior intraparietal sulcus (aIPS), junction between the anterior extent of the intraparietal sulcus and the postcentral sulcus ( Tunik et al., 2005); and (4) ventral premotor cortex (PMv), caudal part of parsopercularis of the inferior frontal gyrus, between the vertical ramus of the lateral fissure and the precentral gyrus ( Davare et al., 2006; 2010). The mean MNI coordinates for all participants were as follows: aSPOC ( x =−17.5 ± 2.5, y =−82.4 ± 3.8, z = 48 ± 5), PMd ( x =−25.6 ± 3.1, y = −0.3.4 ± 4.5, z = 67.5 ± 2.5), aIPS ( x =−50.2 ± 3.7, y =−44.4 ± 5.7, z = 55.8 ± 2.1), and PMv ( x =−57.3 ± 2.2, y = 15.2 ± 3.5, z = 21.1 ± 4.6).
During the experiment, TMS was applied as two pulses spaced 50 ms apart to one of the four brain sites at 120 % of resting motor threshold (RMT). A double pulse was used to extend the virtual lesion effect in order to ensure disruption of processing is coincidental with afferent feedback about the perturbation. Use of double pulses in virtual lesion experiments is common, and prior research has shown that double pulses (40–50 ms ISI) can be used to maintain temporal resolution while maximizing behavioral effects due to the summation of pulse effects on the cortex ( Ellison et al., 2007; Kalla et al., 2008; Luber et al., 2020; O’Shea et al., 2004; Sliwinska et al., 2014). In our experiment the double pulse ensured the duration of the virtual lesion coincided with afferent feedback about the perturbation. On all trials, TMS was automatically triggered by a TTL (transistor-transistor logic, set to 5 V) at the latencies: Early - 100 ms after movement onset and Late - 300 ms after movement onset as described above.
2.7 Finite element (FE) modeling of the electric field (E-field)
FE modeling of the TMS induced electric field was carried out to ensure that stimulation focused on the intended brain regions. Briefly, subject-specific FE head models were produced per subject and TMS-induced E-fields in the brain were computed ( Htet et al., 2019; Saturnino, Madsen, et al., 2019). We used SPM SimNIBS toolbox to create the Head models ( Saturnino, Puonti, et al., 2019). The vector potential of the TMS coil was approximated by small magnetic dipoles ( Thielscher and Kammer, 2004; Windhoff et al., 2013) using the SCIRun problem solving environment and its BrainStimulator toolkit ( Dannhauer et al., 2017; Parker and Johnson, 1995), (see Furmanek et al., 2025 for details).
2.8 Kinematics data processing
Data were analyzed offline using custom Matlab routines (The Mathworks, Natick, MA). Kinematic data were lowpass filtered at 6 Hz with a 4th-order Butterworth filter. Trials were cropped from movement onset (start switch) to moment offset (contact of both thumb and index finger with the object). In VR, the offset was defined as the timestamp when the virtual object was successfully grasped (thumb and index finger markers met the collision detection criteria), ( Furmanek et al., 2021). Data were then interpolated to 100 Hz.
2.9 Statistical analysis
Average aperture and velocity profiles were obtained per subject, per condition, and per brain site. The following kinematic parameters were calculated from these profiles for the full sample: Movement time (MT, time between movement onset and object capture), Peak Aperture (PA, longest distance between index and thumb markers per trial), Time to Peak Aperture (TPA, time point during each trial where PA was achieved), Peak Velocity (PV, the highest value of transport velocity of the wrist marker during a trial) and Time to Peak Velocity (TPV, the time point during each trial where PV was observed).
Repeated Measures (RM) ANOVAs were conducted on the kinematic parameter data to test for differences resulting from the visual perturbation and cortical stimulation across conditions and brain sites. More specifically, we applied 6 × 4 RM ANOVAs with Condition and BrainSite as factors. Furthermore, to test the effects of the Size and Distance perturbations that may not be observable via the kinematic parameters, RM ANOVAs were conducted on the Aperture and Velocity profiles by capturing 5 timepoints at 100 ms intervals following the relevant perturbation. Therefore, for the Early perturbations, the timepoints were 200–600 ms and for the Late perturbations, the timepoints were 400–800 ms. Huyhn-Feldt corrections were used in cases of sphericity violations. Significant ANOVA effects were followed by Bonferroni-Holm post hoc tests to determine the specific differences between conditions.
3 Results
In order to test for non-task related effects of TMS on specific brain regions, a repeated measures ANOVA was conducted with stimulation condition (NoStim, Early, Late) and BrainSite (aIPS, PMv, aSPOC, PMd) as factors for parameters MT, PA, TPA, PV and TPV. Significant main effects of stimulation were observed for most parameters. However, the small effect sizes and non-significant STIM * BrainSite pair-wise comparisons suggest non-specific effects of TMS (
Table 2
3.1 Effects of size perturbation
The kinematic parameters of the effects of the Size perturbation were analyzed to determine the effects of the perturbation at each of the latencies we employed. A 6 × 4 RM ANOVA was conducted with Condition (SNX, LNX, SLEX, SLET, SLLX, SLLT) and BrainSite (AIP, PMv, aSPOC, PMd) as factors. The results are presented in
Table 3.
3.2 Effects of TMS and size perturbation
A 4 × 2 × 2 × 5 Repeated Measures ANOVA with BrainSite (AIP, PMv, aSPOC, PMd), Timing (Early, Late), Stim (TMS, NoTMS), and Timepoint (400–800 ms) was conducted on the grip aperture data. A significant interaction between all factors was observed F (5.57, 44.56)=3.466, p = 0.008, η 2=0.3. Subsequently, a 4 × 2 × 5 RM ANOVA with BrainSite (AIP, PMv, aSPOC, PMd), Stim (TMS, NoTMS), and Timepoint (400–800 ms) was then conducted for each of the Early and Late data. A significant three-way interaction was observed for the Late perturbation only F (9.3, 74.7)=2.53, p = 0.013, η 2=0.024.
In order to look at the effects of TMS on each brain site, 2 × 5 RM ANOVAs were conducted on the data for each brain site. Only aIPS showed significant main effects of STIM F
(1,8)=25.1,
p = 0.002, η
2=0.76 and of Time Point F
(1.2,9.56)=5.25,
p = 0.041, η
2=0.39 and a significant interaction of STIM and Timepoint F
(1.6,12.8)=4.87,
p = 0.03, η
2=0.38. Subsequent Bonferroni-Holm tests were then conducted to identify the time points with significant differences. Timepoints 500 ms through 700 ms were found to be significantly different.
Fig. 3
3.3 Effects of distance perturbation
The kinematic parameters of the effects of the Distance perturbation were analyzed to determine the effects of the perturbation at each of the latencies we employed. A 6 × 4 RM ANOVA was conducted with Condition (SNX, SFX, NFEX, NFET, NFLX, NFLT) and BrainSite (AIP, PMv, aSPOC, PMd). The results are presented in
Table 4
3.4 Effects of TMS and distance perturbation
A 4 × 2 × 2 × 5 Repeated Measures ANOVA with BrainSite (AIP, PMv, aSPOC, PMd), Timing (Early, Late), Stim (TMS, NoTMS) and Timepoint (400–800 ms) was conducted on the transport velocity data. A significant interaction between all factors was observed F (7.17, 57.33)=2.28, p = 0.039, η 2=0.22.
Subsequently, a 4 × 2 × 5 RM ANOVA with BrainSite (AIP, PMv, aSPOC, PMd), Stim (TMS, NoTMS) and Timepoint (400–800 ms) was then conducted for each of the Early and Late data. A significant three-way interaction was observed for the Late perturbation only F (6.96, 55.7)=5.05, p < 0.001, η 2=0.39.
In order to look at the effects of TMS on each brain site, 2 × 5 RM ANOVAs were conducted on the data for each BrainSite. Only PMv showed a significant interaction of STIM and Timepoint F
(4, 32)=6.06,
p < 0.001 η
2=0.43. Subsequent Bonferroni-Holm tests were then conducted to identify the time points with significant differences. Timepoints 600 ms and 700 ms were found to be different (
Fig. 6
4 Discussion
In this study, we took advantage of a VE to produce compelling visual dynamic perturbations of object size and distance. Our previous work has indicated that in spite of the absence of haptic feedback, trained subjects produce well-coordinated reach-to-grasp movements towards the target objects ( Furmanek et al., 2019, 2025).
This experiment involved a direct comparison between the DM and the DL pathways in two types of visual perturbations of the target object: size and distance. Our experimental design included three control conditions, namely, reach to grasp movements to a small object in the near location in the absence of TMS (SNX), plus the same movement with either Early (SNET) or Late (SNLT) TMS. The SNX condition served to frame the effects of both the size (against the large object at the same location, LNX) and the distance (against the same object at the far location, SFX). In the case of SNET and SNLT, these conditions served to control for the effects of both different sizes (against the large object, LNET and LNLT) or different distances (against the same object at the far location, SFET and SFLT).
4.1 Effects of perturbation timing
One of the comparisons in our study involved the contrast between Early (100 msec) and Late (300 msec) perturbation timings. Based on our previous work, we selected those timings to coincide with wrist acceleration and time to peak velocity (TPV) ( Furmanek et al., 2019). Both perturbation timings resulted in effects related to perturbation type: size perturbations caused effects on grasp component parameters like PA and TPA, while distance perturbations affected transport component parameters like PV and TPV. In general, the effects of the Early perturbations were less pronounced than those of the Late perturbations. Nonetheless, in the size perturbation the early change in size of the target object resulted in a larger and delayed peak aperture with respect to the control conditions (SLEX in Fig. 2B). Similarly, in the distance perturbation, the early change in position of the target object caused changes in PV, TPV and MT (NFEX in Fig. 5A). In the case of the Late perturbations, a distinct ‘double-peak’ is observable in the grasping profile during the size perturbation (SLLX in Fig. 2B) and in the transport velocity profile during the distance perturbation (NFLX in Fig. 5A). The size perturbation resulted in a late TPA (the maximum value of the second peak) while the distance perturbation caused a delayed MT as the wrist had to re-accelerate to be able to reach the object at its new position.
Elliott and colleagues ( Elliott et al., 2010, 2017) have proposed that during a goal-directed movement, the guidance processing before TPV is carried out mostly without the involvement of sensory input and through a feedforward comparison to the expected sensory consequences of the movement. Our data support that hypothesis in that Early perturbations resulted in slight modifications of the ongoing movement, while Late perturbations, occurring around TPV produced larger deviations. It has been previously proposed that perturbations of the reach-to-grasp movement that occur in later stages result in a ‘re-programming’ of the movement based on visual input ( Bock and Jungling, 1999; Hesse and Franz, 2009; Paulignan et al., 1991b). This divergence between Early and Late perturbation processing is relevant for our TMS results (see below).
A further interesting point between Early and Late conditions is that our results showed an effect of PMv-TMS on TPA and PV during the early but not the late TMS stimulation relative to the non-perturbed conditions. Disruption effects of TMS on PMv (but not PMd) during non-perturbed grasping are well described ( Davare et al., 2006, 2009; Lega et al., 2020). Indeed, Lega and collaborators observed hand preshaping interference when TMS was applied to PMv early during the movement (50–100 msec after movement onset) but not later, matching our results. It is, therefore, interesting that we did not observe effects of PV-TMS during the Early perturbation condition. It is possible that the perturbation resulted in an updating of the prehension process, delaying PMv’s involvement.
4.2 Effects of TMS
The visual perturbation trials included, in the size conditions, a quick change from a small to a large object with (SLET, SLLT) and without (SLEX, SLLX) TMS at both experimental latencies. In the distance conditions, participants underwent a near to far change in object position also with (NFET, NFLT) and without (NFEX, NFLX) TMS at both experimental latencies. Both of these perturbations resulted in systematic modifications across participants at both latencies, with those following the Late perturbation being more pronounced ( Figs. 2 and 5). Furthermore, TMS to specific brain sites at that latency resulted in significant modifications to particular kinematic parameters during the movement ( Figs. 3 and 6).
More specifically, we observed a significant effect of TMS to aIPS in the compensation of grip aperture for an object size perturbation in the Late timing (300 ms) after movement onset. Relative to the NoTMS condition (SLLX), the TMS condition (SLLT) resulted in a delayed closure of the grip aperture beginning 200 ms after the perturbation. Effects of disruption of aIPS processing on grasp kinematics have been previously reported ( Breveglieri et al., 2023a; Cohen et al., 2009; Tunik et al., 2005). Our data confirmed the causal role of this brain region in the control of the digits during grasping even in the absence of haptic feedback. Furthermore, they support the idea that aIPS, a node of the DL pathway, is critical for digit control during a perturbation of object size, an intrinsic characteristic of the target. Breveglieri and colleagues reported grip aperture disruptions following rTMS to aSPOC in a similar grasping paradigm ( Breveglieri et al., 2023a) at an earlier latency than those observed rTMS to aIPS. In this study, we did not see the effects of TMS to aSPOC in the size perturbation condition. It is possible that differences between our stimulation conditions (five TMS pulses instead of two) may have played a part in this discrepancy. Nonetheless, it is also possible that our grasping task presented fewer constraints on grasping accuracy, which may have resulted in a decreased overall effect on aperture.
The second significant effect of TMS was observed as a modulation of wrist transport velocity in the distance perturbation during the Late timing condition following TMS to PMv. This effect was somewhat unexpected as this premotor region is traditionally associated with the control of grip aperture rather than arm transport ( Buch et al., 2010; Davare et al., 2009; Schettino et al., 2015). Nonetheless, the late distance perturbation resulted in a prominent delayed peak aperture at around 650 ms after movement onset ( Fig. 4A), which corresponds to the timing of the TMS effect on PMv. This result suggests that the double peak observed in the velocity profile occurred in parallel with an adjustment of the grip aperture and may be related to the coordination of the reach and grasp components of the movement.
Our original hypothesis stated that distance perturbation would result in an increased involvement of the DM pathway. For example, in a recent reaching study, Breveglieri and collaborators ( Breveglieri et al., 2023a) found that rTMS to aSPOC shifted arm trajectories when visual targets suddenly changed position. While it is possible that the absence of effects of TMS on either of the nodes of the DM pathway in our study is due to procedural differences, it is important to note that we have recently reported ( Furmanek et al., 2025) an effect of TMS on aSPOC during the compensation to a mechanical perturbation of arm transport. In that study, parallel to the present one, participants underwent a mechanical perturbation in the form of a force produced by a haptic robot along the axis of the reach while producing reach-to-grasp movements to virtual objects. We interpreted that result as a confirmation of the proprioceptive role proposed for aSPOC/V6A ( Breveglieri et al., 2002; Filimon et al., 2009) and of its proposed function as a state estimator integrating proprioception and vision for the compensation of perturbed grasping motions ( Galletti and Fattori, 2018). Relatedly, Galletti and Fattori (2018) also speculated that macaque area V6A could be directly involved in the fast control of the reach-to-grasp movement under time-pressured conditions. Based on this, we expected to see an effect of aSPOC-TMS during the compensation for the distance visual perturbation where the target object changed position as the movement progressed. A plausible explanation for the lack of such an effect is the possibility that the visual perturbation, which was perceived as an instantaneous change in the position of the object rather than a progressive displacement, did not produce the optic flow that aSPOC is purported to detect ( Galletti and Fattori, 2003). One more explanation, as mentioned above, is that the lack of haptic feedback in our VE environment may result in the differential involvement of grasping pathways relative to real-object prehension.
4.3 Limitations
As in our previous studies in a VE, participants in this experiment generated grasping motions that were kinematically comparable to those in physical environments in terms of movement time, aperture and velocity profiles to objects of different sizes ( Castiello et al., 1993; Paulignan et al., 1991b) and at different distances ( Gentilucci et al., 1992; Paulignan et al., 1991a; Zaal and Bootsma, 1998).
Nonetheless, it is important to consider previous research comparing real-world grasping to grasping in virtual reality, particularly as it relates to the absence of haptic feedback. For example, for their study mentioned above, Bock & Jüngling ( Bock and Jungling, 1999) used 2D virtual objects on a screen and, even though they did not find notable differences in most kinematic measures, they did see some differences in final grip aperture. The authors cautioned that lack of haptic feedback during the grasp could affect participants’ behavior. Similar research has suggested the possibility that grasping movements towards non-real objects may rely on different sets of computations ( Freud and Ganel, 2015; Holmes and Heath, 2013; Hosang et al., 2016) et al., 2016) and different brain regions ( Freud and Ganel, 2015; Goodale et al., 1994; Kroliczak et al., 2007). Goodale and collaborators ( 1994, experiment 1) presented real targets of different sizes to a group of neurologically intact participants and asked them to grasp them either directly or 2 s after having removed them from view (pantomimed grasps). They found that pantomimed actions exhibited lower PV, longer MT and smaller PA than movements directed to real objects. Subsequent experiments in the same study involved patient DF, who has ventral visual stream deficits affecting her perception of object shape. DF’s performance in the task suggested to the authors that asking participants to produce pantomimed grasping movements caused them to rely on ventral stream processing, resulting in unnatural prehension. Brain mapping studies, while detecting differences between grasping real and non-real objects, have not detected ventral stream activation ( Freud and Ganel, 2015; Kroliczak et al., 2007).
Also, more recent behavioral work has shown that under certain circumstances, more natural prehension can be observed even in the absence of haptic feedback conditions ( Bingham et al., 2007; Chessa et al., 2019; Furmanek et al., 2019). Nevertheless, it is important to recognize that the absence of haptic feedback may result in kinematic differences such as decreased final aperture ( Bock and Jungling, 1999; Chessa et al., 2019). In our study, however, participants obtained visual feedback of object contact through a collision detection algorithm that has been shown to produce similar final aperture distances relative to grasps towards real objects ( Furmanek et al., 2019). It is possible that this sensory feedback “substitution” may serve to calibrate the movement in a manner that permits natural prehension movements ( Bingham et al., 2007).
One further limitation of the present study is its relatively low sample size. However, it is important to point out that the resource-intensive nature of the study placed practical limits on subject availability. Nonetheless, in order to enhance sensitivity, we selected a repeated measures design. It is not uncommon to use smaller sample sizes in exploratory studies as robust individual responses can still provide meaningful insights.
5 Conclusion
This study confirms the role of the DL pathway in the processing of visual information for the control of the reach-to-grasp movement in the response to visual perturbations and implementation of a corrected motor output. While TMS effects were not observed for the DM pathway, our current results, along with previous work, suggest that combining mechanical and visual perturbations may be a more effective way to assess function in the anterior superior parieto-occipital cortex (aSPOC) and dorsal premotor cortex (PMd). These results provide theoretical support for the development of therapeutic strategies aimed at improving fine motor control in individuals with neurological conditions, such as stroke or cerebral palsy, by targeting specific nodes within the DL pathway. Methodologically, this study also supports the idea that VR environments can be successfully employed for the study of the neural substrates of motor control.
Funding
This work was supported by
CRediT authorship contribution statement
Mariusz P. Furmanek: Writing – original draft, Visualization, Project administration, Methodology, Investigation, Data curation, Conceptualization. Luis F. Schettino: Writing – original draft, Validation, Formal analysis, Data curation, Conceptualization. Mathew Yarossi: Validation, Methodology, Investigation, Conceptualization. Sergei V. Adamovich: Writing – review & editing, Conceptualization. Eugene Tunik: Writing – review & editing, Supervision, Resources.
Declaration of competing interest
The authors have declared no conflict of interest.
Acknowledgements
We acknowledge Alex Huntoon and Samuel Berin for their contributions to the virtual reality platform that enabled this work. We also thank Sambina Anthony for helping with data collection and preparing Fig. 1ABC. We thank Dr. Sumientra Rampersad for helping with FE head models and computing the TMS-induced E-fields in the brain.
Table 1
| CONTROL | |
| SNX - (40) Small object, near, noTMS | |
| SNET - (20) Small object, near, Early TMS | |
| SNLT -(20) Small object, near, Late TMS | |
| SIZE VISUAL PERTURBATION | DISTANCE VISUAL PERTURBATION |
| LNX - (8) Large object, near, noTMS | SFX - (8) Small object, far, noTMS |
| LNET - (4) Large object, near, Early TMS | SFET - (4) Small object, far, Early TMS |
| LNLT- (4) Large Object, near, Late TMS | SFLT - (4) Small object, far, Late TMS |
| SLEX - (8) Small→Large, Early, noTMS | NFEX - (8) Near→Far, Early, noTMS |
| SLET - (8) Small→Large, Early, Early TMS | NFET - (8) Near→Far, Early, Early TMS |
| SLLX - (8) Small→Large, Late, noTMS | NFLX - (8) Near→Far, Late, noTMS |
| SLLT - (8) Small→Large, Late, Late TMS | NFLT - (8) Near→Far, Late, Late TMS |
Table 2
| Variable | STIM | STIM * BrainSite |
| MT (ms) | F 2,16=9.7, p = 0.002*, η 2=0.03 | F 6,48=1.4, p = 0.22, η 2=0.009 |
| PA (cm) | F 2,16=17.4, p < 0.001*, η 2=0.007 | F 6,48=2.28, p = 0.05, η 2=0.002 |
| TPA (ms) | F 2,16=18.7, p < 0.001*, η 2=0.04 | F 6,48=3.4, p = 0.008*, η 2=0.02 |
| PV (cm/s) | F 2,16=5.0, p = 0.02*, η 2=0.001 | F 6,48=3.2, p = 0.008*, η 2=0.002 |
| TPV (ms) | F 2,16=0.9, p = 0.41, η 2=0.007 | F 6,48=1.8, p = 0.103, η 2=0.013 |
Table 3
indicates condition that was different from all others.
indicates a difference between Early and Late conditions.
| Variable | SNX | LNX | SLEX | SLET | SLLX | SLLT |
| MT (ms) | 1155(138) | 1007(104) * | 1118(133) | 1087(99) | 1140(123) | 1126(97) |
| PA (cm) | 8.1(1.4) * | 10.4(1.1) | 10.6(1.1) † | 10.5(1.2) † | 9.3(1.0) | 9.4(1.0) |
| TPA (ms) | 490(67) * | 529(73) | 581(76) † | 586 (66) † | 730 (108) | 685(116) |
| PV (cm/s) | 55.2(5.1) | 55.9(6.4) | 54.7(5.3) | 51.2(14.8) | 55.5(5.1) | 55.3(5.2) |
| TPV (ms) | 297(35) | 288(39) | 293(36) | 292(29) | 299(37) | 298(42) |
Table 4
indicates a condition that was different from all others.
indicates a difference between Early and Late conditions.
| Variable | SNX | SFX | NFEX | NFET | NFLX | NFLT |
| MT (ms) | 1155(138) * | 1225(134) | 1373(141) † | 1299(114) † | 1460(144) | 1446(134) |
| PA (cm) | 8.1(1.4) * | 8.3(1.3) | 8.5(1.5) | 8.3(1.5) | 8.6(1.3) | 8.5(1.3) |
| TPA (ms) | 490(67) * | 567(99) | 556(136) | 565(148) | 654 (150) * | 636(119) * |
| PV (mm/s) | 552(51) | 780(87) * | 607(57) † | 620(53) † | 549(48) | 556(55) |
| TPV (ms) | 297(35.4) | 322(48.1) * | 431(64.6) † | 423(72.8) † | 293(38.9) | 293(32.4) |
© 2025 The Authors