Bench accuracy of a visible-laser and two time-of-flight distance sensor prototypes for near-point measurement on mannequin head phantoms

Purpose: As a step towards a digital alternative to reading a printed ruler scale when measuring thenear point of convergence (NPC) and near point of accommodation (NPA), we evaluated the distanceaccuracy of three N·Examiner prototypes on mannequin head phantoms over 5–50 cm: a Class 2visible-laser meter (V1) and two Class 1 time-of-flight (ToF) devices using the same VL53L4CDsensor (V2, breadboard prototype; V3, integrated hand-held device).Methods: In Experiment 1, V1 and V2 were compared on a matte-black headform. In Experi-ment 2, V1 and V3 were compared on a skin-tone headform mounted on a chin rest with a flat targetplate, using a sliding carriage with a mechanical stop. Five reference distances (5, 10, 20, 30 and50 cm) were each measured 10 times per device with repositioning, plus surface, ambient-light andwarm-up tests at 30 cm. Agreement was assessed by linear regression and Bland–Altman analysis.Results: In Experiment 1, V1 showed a mean bias of +0.1 mm (95% limits of agreement [LoA]−3.7 to +3.9 mm) but no reading at 50 cm on the black surface; V2 showed a proportional positivebias (slope 1.092), from +2.8 mm at 5 cm to +44.4 mm at 50 cm. In Experiment 2, V1 read at alldistances (bias +0.2 mm; LoA −1.7 to +2.1 mm), and V3 showed a small, approximately constantoffset (bias +2.2 mm; LoA −1.3 to +5.7 mm; slope 1.003, 95% confidence interval 1.000–1.006).Surface, ambient light and a 15-minute warm-up changed Experiment 2 mean readings by ≤2 mm.Conclusions: The integrated Class 1 ToF device reproduced phantom reference distances withina few millimetres and without clear evidence of the proportional bias seen with the earlier prototype.As the experiments differed in several factors, the improvement is system-level. Clinical validationin human participants is required.