Due to concerns over patient safety, training via simulation is proposed as a way of supplementing operating theatre experiential learning, especially early in the learning curve [1, 2]. Recent systematic reviews have demonstrated that such training can translate to improved outcomes in the operating theatre [3, 4], although the structure, mode, and design of such training is currently unsystematic [1, 2, 5–8]. The design of training curricula demonstrating validity evidence is therefore a priority for laparoscopic surgery. Such curricula should provide a structure by which an individual can acquire skills to a predetermined level of proficiency, before progressing to more challenging tasks [2].

The measurement and the assessment of technical skills are fundamental to the development of a proficiency-led curriculum. Virtual reality (VR) simulators allow such skills to be practiced in a safe, nonthreatening environment. Furthermore, objective performance assessment (time taken, number of errors, and path length for each hand) is provided immediately, and without the need for monitored supervision [9]. Recent studies have provided validity evidence of these measures, with experienced performers tending to be faster, more accurate, and more efficient in their movement paths than their less experienced counterparts [2, 10–14].

However, while one of the primary difficulties in performing laparoscopic surgery is the translation of two-dimensional video image information to a three-dimensional working area [15, 16], there has been little attention applied to the eye movements and gaze strategies of laparoscopic surgeons [17]. Indeed, the process measures provided by the simulator software reflect the “surgeon-tool” (S–T) interface (tool movement metrics) as opposed to the “surgeon-monitor” (S–M) interface (eye movement / gaze metrics) [18]. A consideration of gaze metrics may provide an important insight into how operators develop the novel spatial relationships required to overcome the perceptual constraints inherent in the laparoscopic environment. Indeed, while tool movement information might provide a proxy indication of the effect of such perceptual disruptions, explicit measures of gaze control provide direct insights into these effects [17, 19].

The aim of the current study was therefore to further our understanding of the psychomotor coordination underlying performance in VR laparoscopic tasks. In addition to comparing expert-novice differences in measures from the separate S–T and S–M interfaces (as did Wilson et al. [19]), a more fine-grained analysis of the spatial and temporal coordination between eye and hand movements, the “quiet eye” [20], was performed for a difficult subcomponent of the task. It was hoped that by examining a measure related to eye-hand coordination, as opposed to hand and eye movements separately, greater insight might be gained into how experienced and novice operators attempt to deal with the perceptual constraints imposed by the laparoscopic environment. In other words, can differences in the way in which eye and hand movements are coordinated to overcome the perceptual constraints in the laparoscopic environment differentiate between experienced and novice laparoscopic operators?

Methods

Participants

A total of 25 surgeons volunteered to take part in the study (mean age = 32.0 years; range = 24–49 years). All participants were right-hand dominant and were classified as novice or experienced laparoscopic surgeons according to the number of laparoscopic procedures they had led. Fifteen novices (6 males, 9 females) had performed fewer than 10 procedures and ten experienced operators (9 males, 1 female) had led more than 60 procedures (range = 60–700). Power calculations using mean and standard deviations from previous studies using similar tasks and groupings [2, 19] suggest that groups should consist of at least eight individuals for a one-tailed test with α = 0.05 and power (1 − β) = 0.8.

Apparatus and task

Testing took place on a LAP Mentor™ (Simbionix USA Corp., Cleveland, OH) VR laparoscopic surgical simulator, based at the Centre for Innovation and Training in Elective Care, Torbay Hospital. The two-handed manoeuvres task from the basic skills training module was used for this study as recent research has suggested that it discriminates between levels of expertise across a range of objective performance and S–T measures [2]. To complete the task the operator must locate balls within a jelly mass and then place them in an endobag. There are three subcomponents to this task requiring accurate psychomotor control: (1) grasping the jelly and manipulating it to expose a ball, (2) grasping a ball, and (3) placing the ball in the endobag (Fig. 1).

Fig. 1
figure 1

Images of the LAP Mentor™ environment showing representative examples of the “two-handed manoeuvres” task: grasping and manipulating the jelly to reveal a ball (A), grasping a ball (B), and dropping a ball in the endobag (C)

Full size image

Participants were fitted with an Applied Science Laboratories Mobile Eye gaze registration system (ASL, Bedford, MA), which measures eye-line of gaze using dark pupil tracking [19]. The system incorporates a pair of lightweight glasses fitted with eye and scene cameras and a set of three LEDs that project harmless near infrared (IR) light onto the eye via a reflective monocle (Fig. 2A). Some of this light is reflected by the cornea (corneal reflection) and appears to the eye camera as a triangle of three dots at a fixed distance from each other. The pupil appears black as light does not exit the inside of the eye, enabling the system to register its position and determine its centre. When the eye turns, the centre of the pupil moves relative to the head, however, the corneal reflection remains in the same position. Therefore, by comparing the vector (angle and distance) between the pupil and the cornea, the eye tracking system can compute the angle at which the eye is pointed (Fig. 2B).

Fig. 2
figure 2

The head-mounted unit from the ASL Mobile Eye gaze registration system (A) and the software environment (B) showing the pupil, the corneal reflection, and the vector line between the two

Full size image

The system also incorporates a recording device (a modified digital video cassette recorder), which combines the two video streams from the eye and scene cameras at 25 Hz. The recorder is attached to a laptop installed with Eyevision (ASL) software; both recorder and laptop were placed on a table to the side of the participant. By teaching the system how the angles calculated by the eye camera relate to the image from the second camera that is viewing the environment (the scene camera), the eye tracker can compute what the eye is pointed at. A circular cursor, representing 1° of visual angle with a 4.5-mm lens, indicating the location of gaze in a video image of the scene (spatial accuracy of ±0.5° visual angle; 0.1° precision), is viewed in real time and recorded for subsequent offline analyses.

Procedure

Participants arrived at the Training Centre individually at prearranged times. They first read an information sheet describing the aims of the study before completing a demographic questionnaire and providing written informed consent of participation. Participants were fitted with the eye tracker and it was calibrated using six visual landmarks on the LAP Mentor display screen. They then performed three consecutive attempts at the two-handed manoeuvres task, as part of a series of activities, before being debriefed and thanked for their participation in the study.

Measures

Complete task

Task performance was assessed in terms of speed (task completion time), and total path length and number of tool movements were chosen from the S–T interface to replicate Aggarwal et al. [2]. Complete task performance and tool process measures were downloaded directly from the LAP Mentor software environment after each trial. The software provided path length and number of movement data for left and right tools and these were aggregated to provide a composite score.

The percentage of time spent fixating on important locations was calculated to provide an overall measure of how the participants attended to the two-dimensional environment (S–M interface) (as in Wilson et al. [19]). The target of interest changes depending on the particular subcomponent of the task being undertaken, i.e., the jelly to be moved in phase 1, the ball to be lifted in phase 2, and the endobag in which to place the ball in phase 3 (Fig. 1). For simpler “pointing” tasks in laparoscopic environments, it has been demonstrated that novices tend to switch gaze between tool and target locations in order to determine the relative position of both, whereas experts adopt a more efficient strategy, fixating almost exclusively on the target location [19, 21]. We were interested in determining if similar gaze strategies were adopted for this grasping task which places greater demands on depth perception than pointing.

The gaze data were analysed in a frame-by-frame manner using GazeTracker (Eye Response Technologies, Charlottesville, VA, USA) video analysis software. A fixation was defined as a gaze of long enough duration to allow information processing (≥120 ms) to a single location (within 1° visual angle). For each subcomponent of the task, areas of interest (“lookzones”) were created around the relevant subtask target (jelly mould, ball, or endobag) and each tool. These were maintained in place by the experimenter as the video progressed at 25 Hz (40 frames a second). The software then provided information regarding the duration and frequency of gazes occurring within each area of interest for the duration of the trial.

Subcomponent task: grasping a ball

In order to provide a more fine-grained analysis of the effects of perceptual constraints on psychomotor control, each attempt at grasping a ball (phase 2) was investigated via video footage analyses. This additional level of analysis sought to answer calls for research examining simulator validity evidence to include outcome measures based on decisive actions during procedures [7]. Ball grasp attempts were selected as they were most sensitive to altered depth perception effects due to the increased degree of accuracy required; the ball diameter is considerably less than either the jelly mould or endobag aperture diameter. The number of attempts required to successfully grasp a ball was defined as the performance measure for this subcomponent, and the quiet eye (QE) was adopted as a specific measure to reflect the spatial and temporal coordination between gaze and motor control [20, 22].

The QE has been shown to underlie higher levels of skill and performance in a wide range of aiming and interceptive skills (see [22] for a review). It is defined as the duration of the final fixation toward a relevant target prior to the execution of the critical phase of movement and has been accepted as a measure of optimal psychomotor control. It is postulated that the QE allows for a period of cognitive preprogramming of movement parameters while minimizing distraction from other environmental or internal cues [20]. The critical movement for this specific task was considered as the arrival of the tool within “2 gaze cursors” distance (i.e., 2°) of the ball. The QE is therefore operationally defined as the final fixation on the ball prior to the arrival of the tool tip within 2° of visual angle of the target.

Analysis

The first trial was considered a familiarization attempt for all participants, providing an insight into the testing protocol while limiting additional learning opportunities prior to testing. Data from the subsequent two trials were averaged to provide a mean value for each variable for each participant to be used for subsequent analyses. The researcher analysing the gaze data was experienced in performing such analyses and blind to the skill levels of the participants to protect against analysis bias.

Shapiro-Wilk tests revealed that all data were normally distributed. Differences between task completion time, total path length (TPL), total number of movements (TNoM), number of attempts required to grasp a ball, and quiet eye (QE) variables for each group were analysed using a series of independent group t tests. Differences in the locations (tools or target) fixated upon were subjected to a mixed-design 2 × 2 ANOVA (group × location), with Bonferroni-adjusted post hoc t tests used to follow up significant interaction effects. All analyses were performed using SPSS 15.0 for Windows (SPSS Inc., Chicago, IL).

Results

Due to the corruption of a data storage device, performance and S–T data from the LAP Mentor for nine participants (2 experts and 7 novices) were lost. While the gaze and subcomponent task data are complete for all 25 participants, the analysis of the LAP Mentor data involves comparisons between eight novices and eight experienced operators. However, these reduced numbers still provide sufficient statistical power [19].

Complete task: performance

Experienced operators completed the task (9 balls) significantly more quickly than novices (t 14 = 3.02, p < 0.010; see Table 1).

Table 1 Mean (±SD) performance and S–T process measures for novice and experienced groups (from LAP Mentor)
Full size table

Complete task: tool movements (S–T interface)

Experienced operators made significantly fewer total movements (t 14 = 3.53, p < 0.005) and had significantly shorter total tool path lengths (t 14 = 3.40, p < 0.005) than novices (see Table 1).

Complete task: gaze strategy (S–M interface)

The ANOVA on the percentage time spent fixating on each gaze location revealed a significant main effect for location (F 1,23 = 27.2, p < 0.001) and no significant main effect for ability level (F 1,23 = 3.4, p = 0.081). These results were qualified by a significant interaction effect (F 1,23 = 13.2, p < 0.005). As Fig. 3 demonstrates, experts spent significantly more time fixating on the relevant target (jelly, ball, or endobag) than their novice counterparts (p < 0.005), while novices spent significantly more time tracking the tools than their expert counterparts (p < 0.005). Novices spent similar amounts of time fixating on the target ball and tracking the tools (p = 0.133), while experts spent significantly more time fixating the target balls compared to tool tracking (p < 0.001).

Fig. 3
figure 3

The percentage of total fixation duration to target (jelly, ball, or endobag) and tool for novice and experienced surgeons (±SEM)

Full size image

Subcomponent task: eye-hand coordination

Experienced operators had significantly longer quiet eye (QE) durations on the target ball (t 23 = 3.49, p < 0.005) and made significantly fewer grasp attempts (t 23 = 2.92, p < 0.010) than novice operators (Table 2).

Table 2 Mean (±SD) quiet eye duration and number of ball grasp attempts during phase 2 of the task, for novice and experienced groups
Full size table

Discussion

The aim of this research was to explore the coordination of eye and tool movements for a laparoscopic training task requiring coordinated and integrated hand movements. This study therefore extends recent research examining separate measures from the S–M (eye movement) and S–T (tool movement) interface for a more basic task [19]. The primary objective was to gain further insight into the strategies used by experienced and novice operators to overcome the perceptual constraints imposed by the laparoscopic environment. As the goal of simulated training is arguably the creation of a “pretrained novice,” prepared for the operating room with reasonably automatic basic psychomotor and visual-spatial laparoscopic skills [6, 10], research examining indices related to these skills is clearly important.

The performance results are in accord with other recent studies and provide support for the discriminatory ability of the two-handed manoeuvres task [2, 12]: Experts were faster in completing the task and had more efficient tool movements than novices (Table 1). While it is often difficult to compare data between studies directly because of differences in procedures and measures reported [7], these results do add to the growing literature base supporting the evidence for validity of some of the LAP Mentor tasks [2, 10–14].

The feedback provided by the LAP Mentor software therefore allows the performance advantage of experts to be characterised and criterion scores for trainee performance to be set [2]. However, it is difficult to get an accurate conceptualisation of the perceptual processes underlying task completion without examining eye movements and indices related to eye-hand coordination. Our basic measure of gaze strategy (where the operators looked during task completion) revealed that experts generally used a target-focused gaze strategy during all three phases of the task, seldom needing to focus on the tool position. In comparison, novices adopted a switching strategy, focusing equally on both target and tool locations [19, 21] (Fig. 3). What does this reveal about differences in the perceptual strategies of both levels of operator?

Skilled psychomotor behaviour involves the ability to predict the consequences of one’s actions and implement mapping rules relating motor and sensory signals [23]. Furthermore, eye movements and the gaze system that controls them play a key role in planning and controlling such precision motor actions [24]. Research examining skilful manual actions (e.g., pointing, grasping) has revealed that goal-directed hand movements are externally driven by target position and do not require visual feedback from the moving hand [25, 26]. Peripheral vision appears to provide sufficient information to refine hand-movement control in this familiar and “natural” environment. This consistent finding can be summarised by the phrase “Keep your eyes on the prize!”

However, the laparoscopic environment is not natural, raising issues of depth perception, elongated tool use, the fulcrum effect, and limited degrees of freedom. Learning therefore involves the adaptation of previously acquired basic sensorimotor rules for manual reaching and grasping [27]. Previous research examining the learning of novel psychomotor tasks has demonstrated the important role of foveal vision and gaze shifts in learning novel mappings between hand motor commands and desirable sensory outcomes [23]. The results we have presented here for laparoscopic surgery reveal similar perceptual processes: Novices switch their attention between targets and tools as they develop the novel mappings required to successfully perform, whereas experienced operators no longer need such exploration and rely on the efficient, target-focused strategy used for less complex environments.

The results for the subcomponent of the task (picking up balls) revealed that it is not only the location of the fixation that is important, but also its timing. Approximately half of the experienced operators’ fixations were not to the target (Fig. 3); however, they utilised a target-focused gaze when it was most needed (the quiet eye, QE). The experienced operators’ QE periods were nearly twice that of the novice operators (Table 2). This steady gaze period appears to have helped guide these precision grasping movements, culminating in fewer unsuccessful attempts (Table 2). These results are therefore supportive of those examining proficiency differences in QE and performance in other motor tasks [22] and suggest that such measures of psychomotor control require further attention in the laparoscopic environment.

The implications of the current research findings extend beyond helping to clarify the processes underlying the skill advantage of expert laparoscopic operators, implying implicitly that these eye movement patterns must be learned and are made more efficient and effective through practice [23]. Given the importance of gaze information in planning and controlling psychomotor skills [24], research attention might profitably be applied to having novices replicate the gaze patterns and psychomotor control of experienced laparoscopic performers [17, 19]. In the sports literature, for example, quiet eye training programmes have successfully expedited novice performance beyond that achieved by providing technical instructions related to movement parameters [28, 29].

We suggest that this advantage may be due to the benefits of learning the skills in an implicit manner, with little conscious awareness of how the skill is executed [30]. By focusing attention externally on relevant targets via gaze control rather than on the movements of the instruments, explicit rules are less likely to be accrued during learning. Masters et al. [31] previously demonstrated that by guiding novices with target information related to suture points, a suturing task could be learned in an implicit fashion. This may be important for subsequent operative performance as the accrual of explicit rules during learning is implicated in skill breakdown under conditions of stress, multitasking, and fatigue [31].

To conclude, the current study has further illuminated how surgeons utilize visual perceptual information to plan and control tool movements in a virtual reality laparoscopic environment. Performance results supported the evidence for validity of the two-handed manoeuvres task from the LAP Mentor basic skills training module. Results also supported previous findings from other domains that have revealed differences in the gaze strategies and psychomotor control (quiet eye) of learners and more experienced performers [20, 23]. Given recent calls for skills training programmes to be based on theoretical frameworks [2, 5, 8], future research should seek to test the utility of gaze training programmes to expedite basic surgical skill learning [17].