Bench validation before hardware selection

1. Confirm the component envelopes. Obtain optical-head, driver, cable and connector CAD for all four channels. Recheck internal overlap, wall thickness, optical-window size, assembly access and heat flow in the actual hollow handle.
2. Measure the optics. Measure lens intrinsics, focus across every mirror, coating reflectance across visible and NIR bands, source divergence and spectral leakage. Calibrate mirror surfaces and camera–projector poses against an independent dimensional target.
3. Test structured-light correspondence. Capture known stepped and crowded dental phantoms. Test wet surfaces, translucency and interreflection. Evaluate decoded completeness and dimensional error separately from the geometric visibility prediction.
4. Separate static and motion performance. Repeat the stationary acquisition first. Then register known small movements, quantify registration drift and count newly reconstructed surface area without duplicates.
5. Validate fluorescence for both goals. With the intended visible-blue source and excitation-rejection filters, measure plaque/biofilm and caries-related contrast independently against appropriate reference assessments. No ordinary PLY colour can supply those labels.
6. Validate NIR as its own modality. Measure reflectance or transillumination with suitable dental samples and known reference findings. Surface ray visibility does not model internal light transport.
7. Repeat assembly and calibration. Test connector reseating, manufacturing tolerances and calibration stability. Report held-out dimensional errors, missing regions and uncertainty, not a single claimed resolution.

The current results are candidates for this bench work. They do not establish clinical performance or a 20–50 µm accuracy specification.