Cinematic rendering in cone-beam computed tomography: A clinical and observer-based evaluation

To compare volume rendering (VR) and cinematic rendering (CR) in cone-beam computed tomography (CBCT)-based three-dimensional (3D) visualisation, focusing on perceived image quality, distance/length estimation accuracy, perceived treatment-planning impact, and practical applicability within clinical workflow.

Fourteen CBCT cases in five categories (implant planning, periodontal, soft-tissue calcifications, endodontic, surgical/maxillofacial) were reconstructed using VR and CR in OsiriX MD. In two sessions (VR first; CR after a 3-week washout), 110 observers (students, general dental practitioners, specialists) rated visual realism, 3D perception, detail sharpness, and perceived impact on anticipated treatment planning on a visual analogue scale and estimated a reference distance in millimetres via open-ended responses. Reconstruction and video export times were recorded.

CR outperformed VR for visual realism, 3D perception, and detail sharpness (+14-19%; all p ≤ 0.007). Distance-estimation error was 2.7% lower with CR, but not statistically significant. Although the perceived impact on anticipated treatment planning favoured CR overall (+14.0%; p > 0.05), gains were statistically significant in surgical/maxillofacial cases (+32%; p < 0.01). Reconstruction was <1 min for both; video export was longer for CR (12-89 min/case) than VR (2-9 min/case).

CR markedly improved perceived image quality over VR. However, its contribution to observer-based measurement accuracy was limited and its perceived impact on anticipated treatment planning was evident, mainly in surgical/maxillofacial cases. Longer export times support selective use alongside VR rather than replacement.

CR may serve as an adjunctive visualisation tool in selected CBCT applications, particularly in surgical/maxillofacial cases, whereas VR remains more practical for routine use.