Effect of Performance Feedback Timing on Motor Learning for a Surgical Training Task

TL;DR

The study tackles how feedback timing affects motor learning in RMIS using a VR ring-on-wire task with multi-sensory cues. It compares real-time (implicit-learning-aligned) feedback, post-task replay (explicit-learning-oriented) feedback, and no feedback, measuring translational and rotational path errors ($TPE$ and $RPE$) and trial time. Results show real-time feedback substantially improves rotational accuracy by the end of training, with replay aiding rotation on long straight segments and no-feedback yielding the least improvement; translational gains are limited to difficult curve regions. The findings support integrating real-time, multi-sensory feedback into RMIS curricula to speed skill acquisition and enhance precision, particularly for complex maneuvers, with implications for ergonomic design and training fidelity.

Abstract

Objective: Robot-assisted minimally invasive surgery (RMIS) has become the gold standard for a variety of surgical procedures, but the optimal method of training surgeons for RMIS is unknown. We hypothesized that real-time, rather than post-task, error feedback would better increase learning speed and reduce errors. Methods: Forty-two surgical novices learned a virtual version of the ring-on-wire task, a canonical task in RMIS training. We investigated the impact of feedback timing with multi-sensory (haptic and visual) cues in three groups: (1) real-time error feedback, (2) trial replay with error feedback, and (3) no error feedback. Results: Participant performance was evaluated based on the accuracy of ring position and orientation during the task. Participants who received real-time feedback outperformed other groups in ring orientation. Additionally, participants who received feedback in replay outperformed participants who did not receive any error feedback on ring orientation during long, straight path sections. There were no significant differences between groups for ring position overall, but participants who received real-time feedback outperformed the other groups in positional accuracy on tightly curved path sections. Conclusion: The addition of real-time haptic and visual error feedback improves learning outcomes in a virtual surgical task over error feedback in replay or no error feedback at all. Significance: This work demonstrates that multi-sensory error feedback delivered in real time leads to better training outcomes as compared to the same feedback delivered after task completion. This novel method of training may enable surgical trainees to develop skills with greater speed and accuracy.

Figures (8)

• Figure 1: Experimental setup. (A) Virtual environment with salient objects labeled. (B) A user interacting with 3D Systems Touch device while the virtual environment is displayed on-screen for the experimenter.
• Figure 2: Sensory augmentations applied to real-time and replay feedback groups during the main trials. (A) Position error: when the ring contacted the wire or went off-path, participants would experience a vibration with constant frequency and amplitude until the ring returned to the wire path. $\text{d}_{\text{vib}}$ represents the direction of vibration in the Y-Z plane for the given position error. (B) Rotation error: when the ring was correctly positioned, i.e., perpendicular to the wire path, it was green. When the ring rotated away from the correct angle, it moved through a gradient from yellow to red depending on magnitude of angle error only, with no directional information about error provided.
• Figure 3: Computed quantities for translational path error (TPE) and rotational path error (RPE). (A) Current position ($T_{n,C}$), desired current position ($T_{n,D}$), and previous desired position ($T_{n-1,D}$). (B) Current pose ($R_{n,C}$), current desired pose ($R_{n,D}$), and angle between the two poses ($\theta_n$).
• Figure 4: Translational path accuracy for first (left) and final (right) trials for sample participant in real-time feedback group. X- and Y-position of ring center are superimposed on true path (light blue). Absolute value of error in Z-position of ring center is indicated by path color, with light-colored sections having highest error.
• Figure 5: Learning curves for all three metrics over the course of the experiment for all three groups ($n =14$ per group, $n = 42$ total). Metrics were computed for each trial for each participant and then averaged across all participants in each group. Lines in first and last section represent baseline and post-test averages; line in middle represents locally estimated scatterplot smoothing curve. Trials marked by stars represent multitasking trials. Black vertical lines after trials 5 and 35 represent introduction and removal of augmentations. Shaded area represents $\pm1$ standard error of the mean.