September 2016
Volume 16, Issue 11
Open Access
Article  |   September 2016
Specific eye–head coordination enhances vision in progressive lens wearers
Author Affiliations & Notes
  • Footnotes
     Commercial relationship: Katharina Rifai and Siegfried Wahl are employed in an industry-on-campus cooperation jointly financed by the Eberhard Karls University of Tuebingen and Carl Zeiss Vision International GmbH.
Journal of Vision September 2016, Vol.16, 5. doi:10.1167/16.11.5
  • Views
  • PDF
  • Share
  • Tools
    • Alerts
      ×
      This feature is available to Subscribers Only
      Sign In or Create an Account ×
    • Get Citation

      Katharina Rifai, Siegfried Wahl; Specific eye–head coordination enhances vision in progressive lens wearers. Journal of Vision 2016;16(11):5. doi: 10.1167/16.11.5.

      Download citation file:


      © ARVO (1962-2015); The Authors (2016-present)

      ×
  • Supplements
Abstract

In uncorrected vision all combinations of eye and head positions are visually equivalent; thus, there is no need for a specific modification of eye–head coordination in young healthy observers. In contrast, the quality of visual input indeed strongly depends on eye position in the majority of healthy elderly drivers, namely in progressive additional lens (PALs) wearers. For a given distance, only specific combinations of eye and head position provide clear vision in a progressive lens wearer. However, although head movements are an integral part of gaze behavior, it is not known if eye–head coordination takes part in the enhancement of visual input in healthy individuals. In the current study we determined changes in eye–head coordination in progressive lens wearers in challenging tasks with high cognitive load, in the situation of driving. During a real-world drive on an urban round track in Stuttgart, gaze movements and head movements were measured in 17 PAL wearers and eye–head coordination was compared to 27 controls with unrestricted vision. Head movement behavior, specific to progressive lens wearers, was determined in head gain and temporal properties of head movements. Furthermore, vertical eye–head coordination was consistent only among PAL wearers. The observed differences in eye–head coordination clearly demonstrate a contribution of head movements in the enhancement of visual input in the healthy human visual system.

Introduction
The modern world provides a diversity of challenges for the visual system. Especially in dynamic and complex situations with high cognitive load, the quality of visual information is a substantial prerequisite for safe and targeted behavior. In this situation gaze behavior is one of the limiting factors for the quality of visual information: Only successful target foveation provides high-resolution visual information. Thus, directing the gaze is a bottleneck for performance in highly dynamic tasks in natural environments. 
Natural scenarios have been analyzed mainly regarding fixation strategies. A variety of studies have analyzed the statistics of fixational patterns, eye movements, and gaze shifts in different contexts, such as everyday tasks (Hayhoe, & Ballard, 2005; Pelz & Canosa, 2001; Pelz, Hayhoe, & Loeber, 2001), walking (Foulsham, Walker, & Kingstone, 2011), consuming movies (Dorr, Martinetz, Gegenfurtner, & Barth, 2010), locomotion (Drew, 1991; Turano, Geruschat, & Baker, 2003), and driving (Lehtonen, Lappi, Koirikivi, & Summala, 2014; Recarte & Nunes, 2000). Also task-dependent changes in gaze statistics have been described there. A comparison between head-free gaze recording and a replay condition further stresses the dominant influence of the visual input (‘t Hart et al., 2009). 
In natural free behavior, the optical quality of the visual input is independent of the head movement executed with accompanying gaze, even though gaze-accompanying head movements follow well-described laws in head-free natural behavior (Einhaeuser et al., 2009; Einhaeuser et al., 2007; Freedman, 2001, 2008; Goossens, & Van Opstal, 1997; Guitton, & Volle, 1987; Hollands, Ziavra, & Bronstein, 2004; Kowler et al., 1992; Land, 1992; Land & Horwood, 1996; Oommen, Smith, & Stahl, 2004; Sağlam, Glasauer, & Lehnen, 2014; Stahl, 1999; Volle & Guitton, 1993; Zangemeister, Jones, & Stark, 1981; Zangemeister & Stark, 1981). Thus, head movements are an integral part of gaze behavior. However, although integral to natural gazing behavior, it is not known if head movements can take part in the enhancement of visual input. 
Progressive additional lenses (PALs), prescribed as accommodative support in presbyopia, a stiffened eye lens in the aged eye, provide a precious role model for the participation of head movements in the enhancement of vision. In presbyopia, the aged, stiffened eye lens is unable to adjust its shape to allow to focus on near objects. A PAL substitutes the accommodation function with gaze-dependent optical lens properties. Whereas there is no need for a specific modification of the head movement component in eye–head coordination in nonpresbyopic observers, in PAL wearers visual input strongly depends on eye position. When gazing through the lower part of the spectacle, near objects appear sharp and in focus, whereas upon gaze through the upper part of the spectacle, far objects appear in focus. Furthermore, due to physical limitations of this lens manufactured as a continuous two surface lens, in oblique directions the optical quality of the PAL degrades. Therefore, major modifications in eye–head coordination are necessary to enhance the quality of visual information. PAL wearers thus give valuable insight in the role of eye–head coordination in the enhancement of visual information. 
Eye–head coordination can efficiently be compared in a dynamic task with high cognitive load, where gaze behavior is a significant performance parameter (Land, 1992). In driving, successful foveation and clear vision are performance sensitive parameters. It contains a constantly challenging visual task demand, randomly assigned in a fast or slow manner, depending on the traffic situation. Thus, eye–head coordination in PAL wearers in the specific situation of driving reveals the role of head movements in the enhancement of visual information. 
In the current study we determined eye–head coordination specific to PAL wearers, thus demonstrating the adaptability of eye–head coordination to visually expressed limitations. 
Methods
The experimental task
Subjects performed a real-world driving task, each of them driving along the same predefined urban round course track in Stuttgart downtown. The subjects were instructed beforehand to follow the directions given by the experimenter, and, naturally, to obey the rules of the German traffic regulations, as well as to drive in a safe and cautious way as they would in their everyday driving activities. Each of the subjects performed the driving course in company of the experimenter who permanently supervised the eye and head tracking recordings from the codriver's seat. The subjects were following the verbal driving instruction of the experimenter. 
The average driving time was 46 min, and the track contained many crossings and turns, as well as traffic lights to induce a high density of opportunities to execute eye–head movements. 
Subjects
Forty-four subjects took part. All subjects gave informed consent in advance of the study. All recordings were performed in accordance with the Declaration of Helsinki. The subjects' average age was 50 years. Twenty-nine of the subjects were male, 15 female. The subjects were divided into two groups, a group of habitual PAL wearers (SD = 53 ± 7 years), and a control group (SD = 44 ± 14 years). All participants had normal or corrected-to-normal vision. All subjects wore their habitual correction, to which they were well adapted. In the PAL group, 17 of the subjects with PALs took part, wearing their habitual progressive correction also during the drive. Twenty-seven subjects were measured in the control group–Seven of those were wearing single-vision lenses during the drive, three were wearing contact lenses, and one was wearing multifocal contact lenses. All subjects were experienced drivers well capable of performing the task. 
The progressive lens as a model system for eye–head coordination
A PAL combines vision correction and accommodation support for presbyopic spectacle lens wearers who suffer from an age-related decay in accommodative abilities. Therefore, PALs provide wearers not only with a correction of their refractive errors, but also with clear vision for every distance. To do so, two lenses are combined in a spatially continuous manner—the corrective lens power in the upper part of the lens and the corrective lens power in addition to the mimicked accommodative power in the lower part of the lens. Progressive lens wearers use the accommodative support gazing downwards on near objects. The geometry of two continuous surfaces of a spectacle lens demands astigmatic errors in the peripheral areas of the lens, leading to a diminishment of optical quality and distortions specifically in the lower peripheral areas of any PAL. This is a mathematical consequence of the Minkwitz–Theorem for two continuous surfaces (Minkwitz, 1963). The exact geometry of a PAL is determined in an iterative mathematical algorithm. These two properties, the increase in power from straight-ahead gaze to downward gaze, as well as the diminishment of optical quality to the periphery act as eye position–specific reward signals for the head movement system. At an advanced state of presbyopia, where the accommodation can be assumed to be fully dysfunctional, the PAL provides clear vision for a specific distance with a predefined head tilt relative to the intended gaze direction only. In addition, the peripheral blur encourages a straight-ahead gaze direction, and therefore a stronger head movement component when gazing at the periphery. Thus, a PAL generates a complex eye position–specific learning signal, rewarding large head movements as well as a specific vertical head position for a specific viewing direction. 
Data acquisition
Throughout the whole drive, binocular gaze (eye in world) and head movements were recorded at 60 Hz by the Smart Eye Pro system (Smart Eye AB, Goteborg, Sweden) in a four camera setup. The tracking accuracy of the system is reported to be 0.5° for head rotation and 0.5° for gaze direction under ideal recording conditions. From the binocular gaze data a cyclopean gaze was determined, which was used in all further analysis. In the data, gaze direction and head direction were defined in a car-centered reference frame. The four cameras were attached to different positions on the dashboard, spanning a minimum recording angle of approximately 160° of visual angle. Each of the subjects was seated in the driving seat, where it was fixed with the seat belt. The subjects were instructed to make themselves familiar with the car, and they were introduced to the eye tracker. Head movements and eye movements were calibrated using the SmartEye calibration routine. The subject was requested to fixate four targets fixed at different positions of the car, spanning the whole recording field. The SmartEye system is a head-free desktop gaze and head tracking system; therefore, the subject was free from any physical contact to the eye tracker, being reminded of its presence only by the existence of the four small cameras as well as two IR-LED arrays, rigidly attached on top of the dashboard. Eye and head movements were continuously recorded during the whole drive. 
Data analysis
From the SmartEye system, gaze and head positions are obtained as unit vectors with horizontal (x), vertical (y), and depth (z) components. In the following, head and gaze movements will be analyzed in this head-centered angle space. Only gaze rotations and head rotations will be considered for analysis. All graphs and numbers will be given in world-centered spatial angles. 
Prior to the analysis erroneous data such as blinks or situations in which either eyes or head were not visible to the tracker were deleted from all datasets. As those data points were defined, either no data was recorded, or head positions or gaze positions clamped to a given value for several hundreds of milliseconds, indicating that the tracker erroneously identified a nonpupil object, which was not moving with the gaze, as the pupil does. 
Eye–head coordination of PAL wearers was analyzed with respect to three main aspects: head gain in eye–head coordination, residual variance in head movements, and temporal properties of eye–head coordination. To do so, each subject's head position was compared to the respective momentary gaze throughout the whole experimental session, regardless of momentary gaze situation or eye movement type, be it a fixation or a saccade. The comparison of these data pairs was done as a linear fit of the function gaze versus head, determining covariation of the two. All eye movement data available from the drive was included, and analysis was done separately for each subject and horizontal and vertical components of the movement. With this approach we thus assume that the free gaze behavior can be approximated by a sequence of static eye–head position combinations, which are determined by the head gain (Volle & Guitton, 1993). This description initially neglects the dynamic phase of a gaze shift, where different delays and time scales of head and eye movements lead to deviations from perfect covariation. Thus, in a second step, a lack-of-fit analysis evaluates deviations from linearity. 
Finally, the origin of any deviation from linearity is quantified separately in a post analysis, assigning differences in eye–head correlation to specific changes in eye–head coordination between the two groups, PAL wearers and controls. For comparison of the two groups, unpaired, two-sided t-tests were carried out. Different properties of the data are compared: gain, correlation (on Fisher's Z), and an analysis of variance separated into fit errors and residual errors. Finally, in a delay analysis head latencies relative to gaze have been compared. 
Results
We measured gaze (eye in world) movements and head movements in 17 PAL wearers (PAL group), and compared them to controls (control group, n = 27). All subjects performed the identical real-world driving task, each of them driving along the same predefined urban round course track in Stuttgart downtown. Subjects were free to move their head and gaze in response to the driving task's needs. 
Figure 1 shows an example trace of gaze and head positions. Indeed, head and gaze move parallel visibly, indicating a high covariation of gaze and head. Head movements show a smoothed behavior in comparison to the gaze movement, confirming the slower time scale of head movements. 
Figure 1
 
Gaze and head movement trace example. Gaze (gray) and head (black) positions in time. Example trace of 125 s taken from one sample subject.
Figure 1
 
Gaze and head movement trace example. Gaze (gray) and head (black) positions in time. Example trace of 125 s taken from one sample subject.
Estimation of head gain
One key descriptor of head movements in eye–head coordination is the head gain. Head gain is determined by a linear fit of the head position relative to the gaze position. Figure 2 shows head direction per gaze direction fit slopes for all subjects, separately for the horizontal and vertical components, with standard errors. Especially horizontally, a strong linear relationship between head and gaze exists. Head position slope is significantly steeper in the PAL group in the horizontal component (mean slope 0.61° head/gaze in PAL group, mean slope 0.51° head/gaze in control group, p < 0.001, unpaired, two-sided t-test). But also vertically the linear fit slope is significantly increased in the PAL group (mean 0.10 in PAL group, 0.02 in control group, p < 0.01, unpaired, two-sided t-test). Thus, a generally stronger head gain was observed in the PAL group. Note furthermore that the slope is close to zero in the control group, indicating that eye and head are not coherently coordinated. 
Figure 2
 
Head gain estimated by a linear fit procedure. (A) Average horizontal linear fit (striped PAL subjects, gray control subjects) with standard errors. (B) Linear fit slopes of horizontal head movements, head component relative to gaze for the PAL group and the control group. (C) Average vertical linear fit (striped PAL group, gray control group) with standard errors. (D) Linear fit slopes of vertical head movements, head component relative to gaze for the PAL group and the control group. A significantly increased slope is visible in the PAL group for both horizontal and vertical movements, indicating an increased head gain in the PAL group.
Figure 2
 
Head gain estimated by a linear fit procedure. (A) Average horizontal linear fit (striped PAL subjects, gray control subjects) with standard errors. (B) Linear fit slopes of horizontal head movements, head component relative to gaze for the PAL group and the control group. (C) Average vertical linear fit (striped PAL group, gray control group) with standard errors. (D) Linear fit slopes of vertical head movements, head component relative to gaze for the PAL group and the control group. A significantly increased slope is visible in the PAL group for both horizontal and vertical movements, indicating an increased head gain in the PAL group.
Lack-of-fit analysis
Figure 3A and C shows head direction and gaze direction for two sample subjects of each group in a correlation contour plot, separately for the horizontal (Figure 3A) and vertical (Figure 3C) components together with a linear fit of the data, shown as black or gray lines, respectively. Figure 3A depicts gaze and head positions in a strong linear relationship, expressed by an oblique orientation of the contour plot of negative (i.e., left) gaze co-occurring with negative (i.e., left) head position, and positive (i.e., right) gaze co-occurring with positive (i.e., right) head position. In Figure 3C, showing the equivalent data for the vertical gaze component, the linear tendency is strongly washed out, but specifically for the PAL group subject, the gaze–head distribution is tilted in the above-described way as well. However, many of the eye–head data pairs strongly deviate from the linear trend. This is expressed in decreases of a perfect correlation, shown in Figure 3B and D
Figure 3
 
Correlation of gaze and head. (A) Correlation contour plot of horizontal head positions relative to gaze positions for two sample subjects (PAL group in black and control group in gray). (B) Group results of the correlation on the head component with the gaze for the PAL group. (C) Correlation contour plot of vertical head positions relative to gaze positions for two sample subjects (PAL group in black and control group in gray). (D) Group results of the correlation on the head component with the gaze for the PAL group and the control group for vertical movements.
Figure 3
 
Correlation of gaze and head. (A) Correlation contour plot of horizontal head positions relative to gaze positions for two sample subjects (PAL group in black and control group in gray). (B) Group results of the correlation on the head component with the gaze for the PAL group. (C) Correlation contour plot of vertical head positions relative to gaze positions for two sample subjects (PAL group in black and control group in gray). (D) Group results of the correlation on the head component with the gaze for the PAL group and the control group for vertical movements.
They might originate from different potential sources (e.g., a nonlinear eye–gaze relationship in eye head coordination). Therefore, the goodness of the linear fit will be compared between the two groups to eliminate a potential confound of nonlinearities in the eye–gaze relationship. 
To do so, systematic deviations are separated in a lack-of-fit analysis, characterized by the fit error (i.e., the difference between the fit value and the local average of the data). The local average of the data was obtained by binning the data in a window of 0.5° for the central 20° of gaze into the respective direction, where the gaze is located most of the time (horizontally, 53% of the time is spent in that area on average; vertically, 83% is spent in this area). As fit error the squared sum of the difference between the local average and the individual data values was defined. The fit error does not differ significantly between the PAL group and the control group, neither horizontally nor vertically (unpaired t-test, phor = 0.34, pver = 0.10). Thus, head movement gain linearity does not differ between the two groups. No major nonlinearities of head position eccentricity relative to gaze eccentricity evolved in the PAL group. 
Residual error
In the next step nonsystematic deviations from linearity are analyzed by the residual error. These nonsystematic deviations might originate from changes in the latency or timing of the head movements of PAL wearers. Alternatively, they might represent gaze-unrelated head movements. In horizontal head movements, the residual error, defined by the variance in head positions for a given gaze position, does not differ significantly between the two groups (unpaired t-test on inverse residuals, p = 0.72, tested for equal variance and normality on the central 20°). Vertically, the average residual error is significantly larger in the control group (unpaired t-test on inverse residuals, p = 0.02, tested for equal variance and normality on the central 20°). Thus, controls on average show a greater variance in vertical head positions. 
Time delay analysis of gaze-accompanying head movements
In the following, it is determined if the larger head variance in controls is originating from potential delays between gaze and head movement. Temporal differences between head and gaze are obvious in Figure 1. Saccades are executed within approximately 50 ms, whereas eye–head movements are mostly finished only 500 ms after an appearing target (e.g., Goossens & Van Opstal 1997; Sağlam, Lehnen, & Glasauer, 2011). In addition, head movements can start delayed in comparison to the saccade (e.g., Herst, Epelboim, & Steinman, 2001; Zangemeister & Stark, 1982). Notably, head movement speed as well as head movement delay can vary. Thus, a change in head movement delay might impact the correlation between head and gaze, which is analyzed in the following. 
Figure 4 shows correlations between head and gaze with an artificially included time delay of various durations between head and gaze for the PAL group in black and the control group in gray. On the x-axis various artificially induced delays between gaze and head are indicated; the curve shows the correlation between gaze and head for each specific artificially induced delay. This correlation curve becomes maximal when the artificially induced time delay compensates for any existent time delay between head and eye movement, and consequently the two movements occur in parallel. The peak of the correlation curve thus indicates the duration of a consistent time delay between gaze and head movement. The correlation is calculated as Fisher's Z of 2-min interval correlations over the whole experimental session. On those only subjects with correlations significantly different from zero were extracted, ensuring that the observed eye–head coordinations are consistently valid over the whole experimental recording. This is valid for the complete group of subjects in the horizontal delay analysis, 12 subjects of the PAL group, and 12 subjects of the control group in the vertical correlation analysis. Figure 4 shows the correlations based on the average Fisher's Z over each whole experimental session. The correlation between gaze and head decays with an increased delay. A positive correlation occurs in both the control and PAL groups for horizontal movements (Figure 4B). For vertical movements, a consistently positive correlation is only visible in the PAL group (Figure 4A). 
Figure 4
 
Delay analysis of gaze-accompanying head movements. (A) Correlation for varying temporal offset between vertical gaze and vertical head for PAL group (black) and control group (gray). (B) Same correlation for horizontal gaze and head. (C) Peak correlation time delay of head for PAL group vertically. (D) Peak correlation time delay of head for PAL group horizontally. (E) Peak correlation time delay of head for control group vertically. (F) Peak correlation time delay of head for control group horizontally.
Figure 4
 
Delay analysis of gaze-accompanying head movements. (A) Correlation for varying temporal offset between vertical gaze and vertical head for PAL group (black) and control group (gray). (B) Same correlation for horizontal gaze and head. (C) Peak correlation time delay of head for PAL group vertically. (D) Peak correlation time delay of head for PAL group horizontally. (E) Peak correlation time delay of head for control group vertically. (F) Peak correlation time delay of head for control group horizontally.
In the bottom rows of Figure 4C through F, the temporal delay estimated by the peak correlation is shown for the PAL group and the control group. No difference in peak correlation time delay between the PAL group and the control group exists for horizontal head movements nor vertical head movements, indicating that there is no measurable change in latency (phor = 0.96, pver = 0.66). Thus, a reduction in overall head movement latency change can be excluded as explanation for the strengthened correlation in the PAL group. 
Taken together, the larger residual error observed in vertical head movements of the control group indicates a less coherent eye–head coordination. 
Results summary
Thus, a larger gain of the head movements was shown horizontally as well as vertically (Figure 2B, D). In addition to this, PAL subjects coherent vertical eye–head coordination (Figure 3D). 
We therefore conclude that eye–head coordination is strengthened in PAL wearers by an increase in head gain as well as by development of tightened vertical coordination between eye and head. 
Discussion
Eye head coordination in progressive lens wearers
Only few descriptions of modifications in eye–head coordination in healthy subjects were shown independently of the external circumstances, demonstrating an inherent, stimulus-independent adjustment of the eye–head coordination system. The present study demonstrates a modified eye–head coordination for healthy subjects wearing progressive lenses under equal stimulus conditions. Until now, only changes in head movements along with changes in saccadic eye movements have been shown in progressive lens wearers in driving simulations, demonstrating a change in exploratory gaze behavior (Chu, Wood, & Collins, 2009). An altered eye–head coordination was demonstrated in reading with progressive lenses, where eye movements can probably be assumed to be stereotyped (Han, Ciuffreda, Selenow, Bauer et al., 2003). Larger horizontal head movements per line and a stronger torsional component in head movements is observed (Han, Ciuffreda, Selenow, Bauer et al., 2003; Han, Ciuffreda, Selenow, & Ali, 2003). Whereas reading is a repetitive task, where line by line, head movements are executed in a similar fashion, in free behavior head movements form a complex pattern of movements, driven by postural changes as well as gaze movements, demanding a complex change in alterations of head movements. 
A strategic enlargement of the head component in the sweep-back saccade could in principle induce the observed behavior. In our study, all amplitudes and directions of head movements were observed, making the application of such a simple strategy impossible. 
Eye–head coordination in the laboratory and in natural behavior
The interplay of head movements with the oculomotor behavior has been studied intensively in terms of eye–head coordination in a laboratory setting (Freedman, 2001, 2008; Goossens & Van Opstal, 1997; Guitton & Volle, 1987; Land, 1992; Oommen, Smith, & Stahl, 2004; Sağlam et al., 2014; Stahl, 1999; Volle & Guitton, 1993; Zangemeister et al., 1981; Zangemeister & Stark, 1981). It is well known that especially larger eye movements are usually accompanied by head movements (Collewijn, Steinman, Erkelens, Kowler, & Van der Steen, 1992; Kowler et al., 1992; Zangemeister, Lehman, & Stark, 1981). Head movements vary, depending on the task (Oommen et al., 2004; Proudlock & Gottlbo, 2007). Also in driving, gaze behavior as well as head movements change in a task-dependent manner (Rahimi, Briggs, & Thom, 1990). A variety of models have been developed describing gaze–head interaction and the control of gaze-accompanying head movements (Boulanger, Galiana, & Guitton, 2012; Daye, Optican, Blohm, & Lefevre, 2014; Hanes & McCollum, 2006; Land, 1992; Mender & Stringer, 2013; Saeb, Weber, & Triesch, 2011; Sağlam, Lehnen, & Glasauer, 2011; Zangemeister, Lehman, & Stark, 1981). Experimentally, under laboratory conditions, a central gaze range is observed, where head movements are rarely executed. For more eccentric gazes, head movements increase approximately linearly (Volle & Guitton, 1993). A piecewise linear relationship for the head component, depending on the initial head position relative to the target, describes isolated eye–head movements well (Volle, & Guitton, 1993). A priori, it is not clear if this relationship applies also in natural behavior. In our head-free self-motivated experiment, we saw continuous head movements. In contrast, the next eye–head movement would only start after the former movement was finished in a typical laboratory experiment to allow an independent analysis. This continuous head movement in self-paced behavior was also observed by Thumser, Oommen, Kofman, and Stahl (2008), who compared it to a stimulus-driven laboratory paradigm. However, they found many similarities between laboratory and natural paradigms. The natural setup showed an additional influence of the exploratory eye range, which is fixed in laboratory paradigms. 
Furthermore, the most obvious difference between the laboratory setting and our natural behavior scenario was a washed-out eye-only region. In our linear fit, an eye-only region would have resulted in a deviation from linearity, but there was no diminishment in head movements for the central region detectable, in agreement with this washed-out eye-only region. This observation is well in agreement with Thumser et al. (2008). It integrates with a description of eye–head coordination depending on the initial head position, which varies in our natural situation (Guitton & Volle, 1987). 
Adjustments in head gain
The most obvious change in head movement behavior observed in the PAL group was a gain increase of horizontal head movements. First evidence for a gain increase in response to visual restrictions in the periphery was observed in two out of three subjects in a more extreme scenario in a study by Stahl, where subjects' head movements were analyzed after wearing a pin-hole (Ceylan, Henriques, Tweed, & Crawford, 2000; Stahl, 2001). The pin hole provided visual information only for large head gains. In our study, visual information was not prevented, but only degraded by optical blur. Nonetheless, consistent increase in head gain was found between PAL wearers and controls. This is, to our knowledge, the first evidence for a change in head gain to enhance vision in natural behavior. 
Vertical eye head coordination in PAL wearers
Vertically, in the control group, head movements did not show a relationship between gaze and head consistent among subjects. Only wearing a progressive lens induced a consistent correlation between gaze and head. This correlation is induced by specific properties of a progressive lens. In a progressive lens, the gradual increase in power with downward gaze favors a specific head position for each object distance and object position. For optimal visual input, a unique mapping exists between eye position and viewing distance. In driving, most of the time, drivers gaze centrally out of the front window observing the traffic, mostly fixating on the street in front of them. Thus, the distance to the street and the objects can be estimated to be in the order of several meters. In this specific situation, the driver uses the far vision zone of the progressive lens with an eye position straight ahead (i.e., a specific eye position). With a unique mapping of eye position and viewing distance, now any vertical gaze position change is translated directly into a head position change, resulting in the observed vertical linear relationship between gaze and head. This made the head position adjustment observable in our data. This is the first clear evidence that a progressive lens enforces a mapping between gaze position and head position, altering eye–head coordination in the complex, stimulus-dependent situation of driving. 
Timing of gaze-accompanying head movements
In the present study the timing of head movements underwent changes. It is known that head movement delays can vary (Zangemeister & Stark, 1982). In the present study head movements were generally barely delayed, probably due to the active status of the subjects during driving. Only vertical anticorrelated head movements, which occurred in some subjects in the control group, showed longer delays up to a few hundred milliseconds. However, vertically a tendency of a faster head movement execution was observed. Although head movements can be executed either way, fast or slow, in eye–head coordination, and also for head movements, a main sequence can be defined (Freedman & Sparks, 2000). However, no detailed reports are known about a learned change in head speed. The consistent correlation between vertical gaze in this study stresses the functional role of the vertical head movements developed in the PAL group. 
Conclusion
The current study demonstrates that head movements participate in the enhancement of vision in human healthy subjects. In unrestricted natural behavior, horizontal head gain is increased. Furthermore vertical gaze-accompanying head movements develop on demand. These results give further insights into the coordination of movements. 
Acknowledgments
The authors thank the Research Institute of Automotive Engineering and Vehicle Engines Stuttgart for their support in data acquisition and Professor Dr. Markus Lappe for his helpful comments on this manuscript. 
Corresponding author: Katharina Rifai. 
Email: katharina.rifai@medizin.uni-tuebingen.de. 
Address: Institute for Ophthalmic Research, University of Tuebingen, Germany. 
References
Boulanger M, Galiana H. L, Guitton D. (2012). Human eye–head gaze shifts preserve their accuracy and spatiotemporal trajectory profiles despite long-duration torque perturbations that assist or oppose head motion. Journal of Neurophysics, 108 (1), 39–56.
Ceylan M, Henriques D. Y. P, Tweed D. B, Crawford J. D. (2000). Task-dependent constraints in motor control: pinhole goggles make the head move like an eye. Journal of Neuroscience, 20 (7), 2719–2730.
Chu B. S, Wood J. M, Collins M. J. (2009). Influence of presbyopic corrections on driving-related eye and head movements. Optometry & Vision Science, 86 (11), E1267–E1275.
Collewijn H, Steinman R. M, Erkelens C. J, Kowler E, Van der Steen J. (1992). Binocular gaze control under free-head conditions. In H. Shimazu & Y. Shinoda (Eds.), Vestibular and brain stem control of eye, head and body movements (pp. 203–220). 220). Basel, Switzerland: Karger.
Daye P. M, Optican L. M, Blohm G, Lefevre P. (2014). Hierarchical control of two-dimensional gaze saccades. Journal of Computer Neuroscience, 36 (3), 355–382.
Dorr M, Martinetz T, Gegenfurtner K. R, Barth E. (2010). Variability of eye movements when viewing dynamic natural scenes. Journal of Vision, 10 (10): 28, 1–17. doi:10.1167/10.10.28. [PubMed] [Article]
Drew T. (1991). Visuomotor coordination in locomotion. Current Opinions in Neurobiology, 1 (4), 652–657.
Einhaeuser W, Moeller G. U, Schumann F, Conradt J, Vockeroth J, Bartl K, Koenig P. (2009). Eye–head coordination during free exploration in human and cat. Annals of the New York Academy of Sciences, 1164 (1), 353–366.
Einhaeuser W, Schumann F, Bardins S, Bartl K, Boening G, Schneider E, Koenig P. (2007). Human eye-head co-ordination in natural exploration. Network: Computation in Neural Systems, 18 (3), 267–297.
Foulsham T, Walker E, Kingstone A. (2011). The where, what and when of gaze allocation in the lab and the natural environment. Vision Research, 51 (17), 1920–1931.
Freedman E. G. (2001). Interactions between eye and head control signals can account for movement kinematics. Biological Cybernetics, 84 (6), 453–462.
Freedman E. G. (2008). Coordination of the eyes and head during visual orienting. Experimental Brain Research, 190 (4), 369–387.
Freedman E. G, Sparks D. L. (2000). Coordination of the eyes and head: Movement kinematics. Experimental Brain Research, 131 (1), 22–32.
Goossens H. H, Van Opstal A. J. (1997). Human eye-head coordination in two dimensions under different sensorimotor conditions. Experimental Brain Research, 114 (3), 542–560.
Guitton D, Volle M. (1987). Gaze control in humans: Eye–head coordination during orienting movements to targets within and beyond the oculomotor range. Journal of Neurophysics, 58 (3), 427–459.
Han Y, Ciuffreda K. J, Selenow A, Ali S. R. (2003). Dynamic interactions of eye and head movements when reading with single-vision and progressive lenses in a simulated computer-based environment. Investigative Ophthalmology & Vision Science, 44 (4), 1534–1545. 1545. [PubMed] [Article]
Han Y, Ciuffreda K. J, Selenow A, Bauer E, Ali S. R, Spencer W. (2003). Static aspects of eye and head movements during reading in a simulated computer-based environment with single-vision and progressive lenses. Investigative Ophthalmology & Vision Science, 44 (1), 145–153. 153. [PubMed] [Article]
Hanes D. A, McCollum G. (2006). Variables contributing to the coordination of rapid eye/head gaze shifts. Biological Cybernetics, 94 (4), 300–324.
Hayhoe M, Ballard D. (2005). Eye movements in natural behavior. Trends in Cognitive Science, 9 (4), 188–194.
Herst A. N, Epelboim J, Steinman R. M. (2001). Temporal coordination of the human head and eye during a natural sequential tapping task. Vision Research, 41 (25), 3307–3319.
Hollands M. A, Ziavra N. V, Bronstein A. M. (2004). A new paradigm to investigate the roles of head and eye movements in the coordination of whole-body movements. Experimental Brain Research, 154 (2), 261–266.
Kowler E, Pizlo Z, Zhu G, Erkelens C. J, Steinman R. M, Collewijn H. (1992). Coordination of head and eyes during the performance of natural (and unnatural) visual tasks. In A. Berthoz, W. Graf, & P. P. Vidal (Eds.), The head-neck sensory motor system (pp. 419–426). 426). Oxford, UK: Oxford University Press.
Land M. F. (1992). Predictable eye–head coordination during driving. Nature, 359 (6393), 318.
Land M, Horwood J. (1996). The relations between head and eye movements during driving. Vision in Vehicles, 5, 153–160.
Lehtonen E, Lappi O, Koirikivi I, Summala H. (2014). Effect of driving experience on anticipatory look-ahead fixations in real curve driving. Accident Analysis & Prevention, 70, 195–208.
Mender B. M, Stringer S. M. (2013). A model of self-organizing head-centered visual responses in primate parietal areas. PloS One, 8 (12), e81406.
Minkwitz G. (1963). Über den Flächenastigmatismus bei gewissen symmetrischen Asphären [Translation: On surface astigmatism of specific symmetric aspheric lenses]. Journal of Modern Optics, 10 (3), 223–227.
Oommen B. S, Smith R. M, Stahl J. S. (2004). The influence of future gaze orientation upon eye-head coupling during saccades. Experimental Brain Research, 155 (1), 9–18.
Pelz J. B, Canosa R. (2001). Oculomotor behavior and perceptual strategies in complex tasks. Vision Research, 41 (25), 3587–3596.
Pelz J, Hayhoe M, Loeber R. (2001). The coordination of eye, head, and hand movements in a natural task. Experimental Brain Research, 139 (3), 266–277.
Proudlock F. A, Gottlob I. (2007). Physiology and pathology of eye–head coordination. Progress in Retinal and Eye Research, 26 (5), 486–515.
Rahimi M, Briggs R. P, Thom D. R. (1990). A field evaluation of driver eye and head movement strategies toward environmental targets and distractors. Applied Ergonomics, 21 (4), 267–274.
Recarte M. A, Nunes L. M. (2000). Effects of verbal and spatial-imagery tasks on eye fixations while driving. Journal of Experimental Psychology: Applied, 6 (1), 31.
Saeb S, Weber C, Triesch J. (2011). Learning the optimal control of coordinated eye and head movements. Plos Computational Biology, 7 (11), e1002253.
Sağlam M, Glasauer S, Lehnen N. (2014). Vestibular and cerebellar contribution to gaze optimality. Brain, 137 (4), 1080–1094.
Sağlam M, Lehnen N, Glasauer S. (2011). Optimal control of natural eye-head movements minimizes the impact of noise. Journal of Neuroscience, 31 (45), 16185–16193.
Stahl J. S. (1999). Amplitude of human head movements associated with horizontal saccades. Experimental Brain Research, 126 (1), 41–54.
Stahl J. S. (2001). Adaptive plasticity of head movement propensity. Experimental Brain Research, 139 (2), 201–208.
'tHart M. B. , Vockeroth J, Schumann F, Bartl K, Schneider E, Koenig P, Einhäuser W. (2009). Gaze allocation in natural stimuli: Comparing free exploration to head-fixed viewing conditions. Visual Cognition, 17 (6–7), 1132–1158.
Thumser Z. C, Oommen B. S, Kofman I. S, Stahl J. S. (2008). Idiosyncratic variations in eye–head coupling observed in the laboratory also manifest during spontaneous behavior in a natural setting. Experimental Brain Research, 191 (4), 419–434.
Turano K. A, Geruschat D. R, Baker F. H. (2003). Oculomotor strategies for the direction of gaze tested with a real-world activity. Vision Research, 43 (3), 333–346.
Volle M, Guitton D. (1993). Human gaze shifts in which head and eyes are not initially aligned. Experimental Brain Research, 94 (3), 463–470.
Zangemeister W. H, Jones A, Stark L. (1981). Dynamics of head movement trajectories: main sequence relationship. Experimental Neurology, 71 (1), 76–91.
Zangemeister W. H, Lehman S, Stark L. (1981). Simulation of head movement trajectories: model and fit to main sequence. Biological Cybernetics, 41 (1), 19–32.
Zangemeister W. H, Stark L. (1981). Active head rotations and eye-head coordination*. Annals of the New York Academy of Sciences, 374 (1), 540–559.
Zangemeister W. H, Stark L. (1982). Types of gaze movement: variable interactions of eye and head movements. Experimental Neurology, 77 (3), 563–577.
Figure 1
 
Gaze and head movement trace example. Gaze (gray) and head (black) positions in time. Example trace of 125 s taken from one sample subject.
Figure 1
 
Gaze and head movement trace example. Gaze (gray) and head (black) positions in time. Example trace of 125 s taken from one sample subject.
Figure 2
 
Head gain estimated by a linear fit procedure. (A) Average horizontal linear fit (striped PAL subjects, gray control subjects) with standard errors. (B) Linear fit slopes of horizontal head movements, head component relative to gaze for the PAL group and the control group. (C) Average vertical linear fit (striped PAL group, gray control group) with standard errors. (D) Linear fit slopes of vertical head movements, head component relative to gaze for the PAL group and the control group. A significantly increased slope is visible in the PAL group for both horizontal and vertical movements, indicating an increased head gain in the PAL group.
Figure 2
 
Head gain estimated by a linear fit procedure. (A) Average horizontal linear fit (striped PAL subjects, gray control subjects) with standard errors. (B) Linear fit slopes of horizontal head movements, head component relative to gaze for the PAL group and the control group. (C) Average vertical linear fit (striped PAL group, gray control group) with standard errors. (D) Linear fit slopes of vertical head movements, head component relative to gaze for the PAL group and the control group. A significantly increased slope is visible in the PAL group for both horizontal and vertical movements, indicating an increased head gain in the PAL group.
Figure 3
 
Correlation of gaze and head. (A) Correlation contour plot of horizontal head positions relative to gaze positions for two sample subjects (PAL group in black and control group in gray). (B) Group results of the correlation on the head component with the gaze for the PAL group. (C) Correlation contour plot of vertical head positions relative to gaze positions for two sample subjects (PAL group in black and control group in gray). (D) Group results of the correlation on the head component with the gaze for the PAL group and the control group for vertical movements.
Figure 3
 
Correlation of gaze and head. (A) Correlation contour plot of horizontal head positions relative to gaze positions for two sample subjects (PAL group in black and control group in gray). (B) Group results of the correlation on the head component with the gaze for the PAL group. (C) Correlation contour plot of vertical head positions relative to gaze positions for two sample subjects (PAL group in black and control group in gray). (D) Group results of the correlation on the head component with the gaze for the PAL group and the control group for vertical movements.
Figure 4
 
Delay analysis of gaze-accompanying head movements. (A) Correlation for varying temporal offset between vertical gaze and vertical head for PAL group (black) and control group (gray). (B) Same correlation for horizontal gaze and head. (C) Peak correlation time delay of head for PAL group vertically. (D) Peak correlation time delay of head for PAL group horizontally. (E) Peak correlation time delay of head for control group vertically. (F) Peak correlation time delay of head for control group horizontally.
Figure 4
 
Delay analysis of gaze-accompanying head movements. (A) Correlation for varying temporal offset between vertical gaze and vertical head for PAL group (black) and control group (gray). (B) Same correlation for horizontal gaze and head. (C) Peak correlation time delay of head for PAL group vertically. (D) Peak correlation time delay of head for PAL group horizontally. (E) Peak correlation time delay of head for control group vertically. (F) Peak correlation time delay of head for control group horizontally.
×
×

This PDF is available to Subscribers Only

Sign in or purchase a subscription to access this content. ×

You must be signed into an individual account to use this feature.

×