JMIR Publications

ABSTRACT

Background:

Augmented reality (AR) head mounted displays (HMDs) could overcome the spatial dissociation between medical imaging and the surgical fields. This may be particularly important in anatomically dense regions such as the head and neck. Although many HMDs offer markerless inside-out tracking at a fraction of the price of navigation systems, their overlay accuracy is limited. This limit is especially critical in the superimposition (SI) of holography onto the surgical field. The virtual twin (VT), where holography is visualized directly adjacent to the surgical field, could be a viable alternative. However, its performance is still unclear.

Objective:

To compare the accuracy and efficiency of the two visualization modalities SI and VT for anatomical localization in the head and neck region

Methods:

In a randomized crossover trial to compare two AR visualization modalities (SI and VT), 38 participants used a HoloLens 2 (HL2) to localize nerve exit points, inferior alveolar nerves, and salivary glands on head phantoms. Their performance was evaluated in terms of accuracy, workload, and user experience.

Results:

SI achieved better total accuracy with 14.4 ± 4.2 mm and relative accuracy with 3.4 ± 2.2 mm in point localization, compared to VT with 15.8 ± 5.5 mm and 6.0 ± 5.0 mm, respectively (total: p = 0.003; relative: p < 0.001). In linear structures, accuracy performance was comparable between SI (ASD: 23.4 ± 4.1 mm; HD 31.5 ± 7.8 mm) and VT (ASD: 23.0 ± 4.5 mm; HD 31.0 ± 7.5 mm) (p > 0.05). However, SI showed significantly higher deviation than VT in volume-based structure localization accuracy (SI ASD: 37.1 ± 13.8 mm vs. VT ASD: 34.1 ± 14.2 mm, p = 0.011; SI HD: 52.0 ± 16.8 mm vs. VT HD: 49.1 ± 15.8 mm, p = 0.034). Subjects were faster with SI (p = 0.017), while workload NASA-TLX scores didn’t show a difference (p > 0.05).

Conclusions:

Given that SI did not clearly outperform VT under overlayed soft tissue and viewing challenges, VT remains a viable alternative for certain surgical scenarios in tasks where high accuracy is not required. To address the challenges of AR visualization in the future, viewing angles could be guided and targets could be linked to the skin surface. Clinical Trial: German Clinical Trial Register (DRKS00032835)