Fig 1.
A: Example of acupuncture manipulation education system (AMES) and two different types of visual feedbacks: concurrent and post-trial visual feedback. B: Experimental setup and procedure.
The experiment comprised two sessions: simple and complicated lifting/thrusting movements. Each session consisted of (1) a pre-training test, (2) training across eight trials, and (3) a post-training test. During the pre- and post-training tests, participants performed acupuncture needle manipulations without any visual information, and aimed to mimic the motion template as closely as possible (+). A simulation of their own motion (green line) overlapped the motion template (red line) in the concurrent visual feedback (VF) group; the motion information (green line) of the post-trial VF group was presented, along with the motion template (red line), immediately subsequent to each trial.
Fig 2.
Data processing and statistical analysis.
Identified sampled motion unit (a). Slicing and collection sampled motion units (b). Normalization to resample the observed data to a specific number (i.e., 50) and calculated magnitude error (c). Normalization with rescaling of the lift/thrusting amplitude to between 0 and 1 (d). Calculated shape and mean squared error (e).
Fig 3.
A: Estimated regression curve for the motion patterns of each participant during acupuncture needle manipulation (red dot: pre-training test, blue dot: post-training test). Mean response curve of the motion patterns of each group (red line: pre-training test, blue line: post-training test, black line: a motion template for the lifting/thrusting movement). Both the concurrent and post-trial VF groups exhibited a greater number of deviations from the motion template during the pre-training test (red line) and fewer deviations during the post-training test (blue line) in session 2. B: Both the concurrent and post-trial VF groups exhibited greater motion pattern improvements during the complicated, but not the simple, lifting/thrusting session.
Fig 4.
A: In the magnitude error analysis, both the concurrent and post-trial VF groups exhibited greater error reductions during the simple lifting/thrusting session. B: During the training period, the concurrent VF group exhibited reduced error magnitudes across all trials, while the post-trial VF group exhibited greater error magnitudes during the initial trials, which gradually decreased in later trials.
Fig 5.
The characteristics of both conditions (concurrent and post-trial VF) are summarized.
The modality, latent process, and results of each condition are included to facilitate understanding of each visual feedback condition.