Effect of binocular disparity on learning anatomy with stereoscopic augmented reality visualization: A double center randomized controlled trial

Abstract

Binocular disparity provides one of the important depth cues within stereoscopic three-dimensional (3D) visualization technology. However, there is limited research on its effect on learning within a 3D augmented reality (AR) environment. This study evaluated the effect of binocular disparity on the acquisition of anatomical knowledge and perceived cognitive load in relation to visual-spatial abilities. In a double-center randomized controlled trial, first-year (bio)medical undergraduates studied lower extremity anatomy in an interactive 3D AR environment either with a stereoscopic 3D view (n = 32) or monoscopic 3D view (n = 34). Visual-spatial abilities were tested with a mental rotation test. Anatomical knowledge was assessed by a validated 30-item written test and 30-item specimen test. Cognitive load was measured by the NASA-TLX questionnaire. Students in the stereoscopic 3D and monoscopic 3D groups performed equally well in terms of percentage correct answers (written test: 47.9 ± 15.8 vs. 49.1 ± 18.3; P = 0.635; specimen test: 43.0 ± 17.9 vs. 46.3 ± 15.1; P = 0.429), and perceived cognitive load scores (6.2 ± 1.0 vs. 6.2 ± 1.3; P = 0.992). Regardless of intervention, visual-spatial abilities were positively associated with the specimen test scores (η² = 0.13, P = 0.003), perceived representativeness of the anatomy test questions (P = 0.010) and subjective improvement in anatomy knowledge (P < 0.001). In conclusion, binocular disparity does not improve learning anatomy. Motion parallax should be considered as another important depth cue that contributes to depth perception during learning in a stereoscopic 3D AR environment.

Keywords: anatomical education; stereoscopic three-dimensional technology; visual-spatial abilities.

FIGURE 1

Flowchart of study design. 3D, three‐dimensional; LUMC, Leiden University Medical Center; n, number of participants; Radboudumc, Radboud University Medical Center

FIGURE 2

Differences in overall mean percentages correct answers on the (A) written knowledge test, (B) specimen knowledge test and (C) cognitive load test between stereoscopic 3D view (n = 32) and monoscopic 3D view (n = 34) groups. 3D, three‐dimensional; ^anot statistically significant differences

FIGURE 3

Relationship between mental rotation test (MRT) scores and specimen test scores. A regression analysis graph illustrating a positive association between visual‐spatial abilities and specimen test scores. MRT, mental rotation test